Jump to content

Wikidata:Requests for comment/Notability policy reform

From Wikidata

Wikidata’s current Notability Policy was drafted in 2013. Since then, the project has grown and the world has changed. We understand Wikidata, what it can do and what its place in the Linked Open Data web is, better. All of this led to discussions about what data should be in Wikidata, what should be in the larger Wikibase Ecosystem and what should maybe go somewhere else entirely. This RfC is meant to create an improved Notability Policy.

Context

[edit]

What has changed since the initial drafting of the policy

[edit]
  • Wikidata has grown to 120 Million Items and 16.000 active editors
  • Wikidata has reached social and technical scaling issues (see Wikidata:Requests for comment/Mass-editing policy#Context for additional information)
  • The Wikibase Ecosystem has matured, with Wikibase Cloud and Wikibase Suite now being viable alternatives for hosting knowledge graphs that are tightly interconnected with Wikidata

What can we learn from other projects

[edit]

Looking at other Notability Policies of other Wikimedia Projects, a few things stood out that we might want to take into account:

  • Wikimedia Commons: It clearly calls out the aim of Commons to ground the policy. It also specifically mentions some content not being a good fit for Commons because there are other projects that handle it. It talks very clearly about requiring use of the files or at the very least requires files to be realistically useful for educational purposes.
  • German-language Wikipedia: It has de:Wikipedia:Relevanzcheck where people can ask before creating a new article if it would meet notability criteria.
  • There is a general theme of requiring reliable, independent, secondary sources.
  • There are projects that define notability generally for all topics (like Wikidata does now), while others go deeper into specific notability criteria for specific topics.

The benefits and problems of the current policy

[edit]

Based on previous discussions there are a number of positive and negative points to the current policy.

Benefits:

  • Small number of criteria: makes it easier to understand and to remember
  • Wiggle-room for marginalized knowledge and more: There is benefit to the current vagueness. It gives admins room to make judgement calls.

Problems:

  • Differing interpretations of criteria 2: Some people see it as a carte blanche (everything can have an Item); others interpret it much more strictly. Admins report not being able to be consistent even within their own decisions. It is also harder to objectively evaluate than criteria 1 and 3.
  • Level of consideration is mismatched: The policy works at the level of individual Items but issues are also (even more so) caused by collection of Items even if every individual one of them is notable. We don’t have good ways to draw boundaries about what from a class we want in Wikidata and what we don’t.
  • Enforcement burden: Admins spend a lot of time arguing over the interpretation of and enforcing the policy. People not acting in the best interest of the project create Items to promote themselves, their business, etc, causing more clean-up work for admins.
  • Detachment from reuse: Reuse is not really considered in the criteria but reuse is at the core of what Wikidata is about. Additionally, Items get deleted as not notable despite being used in other projects.

Round 1: Questions to determine changes to current policy

[edit]

Following are a number of questions that need your input to help draft the new policy.

Questions and discussion that was used to draft the new policy

We should require reuse or at least the very real potential for reuse

[edit]

Votes and comments:

Take-away for policy drafting:

@Mariana Fossatti (WK?): Please use the standard ~~~~ signature so that your comments have a timestamp and the reply link. ArthurPSmith (talk) 18:09, 23 January 2026 (UTC)reply

We should consider the size of the complete data set when making decisions about individual Items

[edit]

Votes and comments:

  • I think this is intended for cases like "all known stars" or "all the scientific articles ever published", which presently could not go into Wikidata at least for technical reasons; in this sense, I think it makes sense considering this criterium in a future change of the notability policy; however, an item could be considered part of many different datasets, each of a different size (like, for a scientif article: all the articles published in the same volume, or in all the volumes of the same journal, or in all the volumes of the journals treating this topic in language X, or in all journals treating this topic in all languages etc.), so we should be very careful on applying this IMHO. --Epìdosis 18:14, 30 December 2025 (UTC)reply
I think you have a point here, but I don't think the carefulness is a hindrance. We just need to be explicit about which set we are talking about and be clear that we don't fall into a scope creep. Ainali (talk) 10:02, 31 December 2025 (UTC)reply
  • I think this clause related (somehow) to the reusability clause above. One issue popped in my mind is "when theres only a subset of the data in the Wikidata, what will persuade us to use that instead the complete data in its original source? Does Wikidata will serve as 'short preview' to the complete data set?" but I do agree  Support if we must consider the size of data. —Yamato Shiya大和 士也 (Talk • Contribs) 12:16, 31 December 2025 (UTC)reply
  •  Oppose We should fix the wikibase-backend to scale as the community wants instead of limiting creation of new items. See Wikidata:Requests_for_comment/Notability_policy_reform#Is_there_anything_we_missed? below for details about the work on a new wikibase-backend that scales beyond 1bn+ entities.--So9q (talk) 22:29, 1 January 2026 (UTC)reply
     Support I agee. Given the current policy lasted over 10 years, making decisions over this scale based on current capacity limitations is the wrong reason. GrimRob (talk) 18:08, 30 January 2026 (UTC)reply
  •  Comment I think this is somewhat redundant - the word 'notability' implies that the entity in question stands out from others of its type in some way or other. So inclusion can't be just accepted based on being a member of a (very large) group, there must be some narrower criteria involved. Perhaps the notability text should say something like "member of a reasonably sized class, not one of billions"? ArthurPSmith (talk) 16:48, 3 January 2026 (UTC)reply
    I think WD has a clear goal to be comprehensive, more so than an encyclopedia, and we'd not want to use a notability argument to cherry pick members of a class, when the minor members have a structural need to complete the set. But its completeness based on extending current data sideways, so we can say to users that we can only accommodate 200m items with current technology, and our roadmap to that anticipates your item won't make the cut. The problem is our uneven coverage of knowledge, to be hard-nosed we should offload swathes of items already in the system to other projects, but that's technically and socially difficult, and until then someone can always say "but you let X in, why not my Y". And while we can rank notability within a field, its much harder across fields. Our graph system is best for spotty, complicated data sets, uniform ones like stars catalogues are better elsewhere. Vicarage (talk) 19:35, 3 January 2026 (UTC)reply
  •  Oppose Jerimee (talk) 20:11, 15 January 2026 (UTC)reply

Take-away for policy drafting:

  • This should be carefully considered in light of Wikidata's scope and main goals. One of them is to serve as a general knowledge base, and currently, this is one of the most critical aspects of Wikidata, since the extensive use made by third-party apps and AI models. A general database in which items are highly curated in order to keep only “notable” items from a data set faces the risk of determining an exclusionary selection with potential biases. The risk is greater than on Wikipedia (where this dynamic already exists and creates well-documented biases). An incomplete database can lead to even more biases when it is reused for research and other purposes. Furthermore, bias can spread beyond Wikidata to study results, apps, and more. P.S Lydia Pintscher (WMDE) I know that I'm coming late to the discussion. I'm responding in general to the answers, not directly to your take-away comments (which I really appreciate to read). Mariana Fossatti (WK?) (talk) 17:05, 23 January 2026 (UTC)reply

We should change the wording of criteria 2 to say “can be described” to “is described”

[edit]

Votes and comments:

Take-away for policy drafting:

  • Further clarification is needed for this proposal. What is suggested as a small change in wording seems to play a major role in solving an important problem. But what are we solving for? Why is this wording the required solution? In any case, some proposals previously shared, such as limiting notability only to items that have identifiers, will narrow Wikidata scope drastically, putting at risk the capacity of editors to cover marginalized topics. Often, editors who cover such topics can’t find the item's external IDs in mainstream databases, and many identifier properties, especially from Global South institutions, are not yet available on Wikidata. Also, in order to make good descriptions in marginalized topics, we have to navigate across invisibilized, fragile and fragmented sources that are spread across multiple materials. It is fair –and important– to ask for references, but it is also important to admit items with essential statements based on initial research, keeping them open to further improvement. We can do better in signaling where more references and maintenance are needed, in order to refine and improve the existing data (automation could be helpful here). Mariana Fossatti (WK?) (talk) 17:07, 23 January 2026 (UTC)reply
  • I agree withc Mariana Fossatti (WK?). In the Global South, and in countries such as Colombia, there is not a strong culture of open data. Finding databases with unique identifiers that can be linked to Wikidata items is often extremely difficult. As a result, in order to maintain our knowledge on Wikidata in a way that meets notability requirements, we often rely more heavily on references. At the same time, some topics are only beginning to be studied or remain contested within the historical memory of our countries. In these cases, important references do exist, but they may be at risk of censorship or deletion, or they may simply come from emerging areas of research that are only now beginning to be documented. For this reason, I would ask that this issue be considered very carefully. Otherwise, a significant amount of knowledge from the Global South—knowledge that is important to our communities and needs to circulate and be preserved—could be left out of a project as important as Wikidata. AmiGueko (talk) 21:56, 14 July 2026 (UTC)reply

Accompanying text: We should require new Items to have at least the statements required to establish notability

[edit]

Votes and comments:

  • Harsh given we have so many items already that don't have that. And in a world of bots and AI, many items might be created in anticipation of waves of bot updates that could be some time away. Scurrying around at creation to find mentions elsewhere would encourage random facts being added, like references being to obscure newspaper articles some are so fond of. Better to consistently back-fill across a class from notable sources. Vicarage (talk) 16:56, 30 December 2025 (UTC)reply
  • I agree with Vicarage that this could backfire, but IMHO the pros are more than the cons: after an item has been created, and some time has passed (say one day), the creator should have added to it sufficient data to demonstrate that it is notable. --Epìdosis 18:22, 30 December 2025 (UTC)reply
  •  Support But, admins should not delete empty items minutes after they are created (time should be given for the creator to fill the item). Ternera (talk) 19:30, 30 December 2025 (UTC)reply
    Agreed. For what it's worth, English Wikipedia's New Page Patrollers are required to wait until an hour after a page has been created before nominating something for deletion unless there are serious content issues, like BLP privacy issues. Perhaps something similar would make sense here. Mcampany (talk) 03:07, 31 December 2025 (UTC)reply
    I feel that it is reasonable to expect editors to add a claim within fifteen minutes of item creation, but I normally do not delete empty items until an hour after the last edit. Bovlb (talk) 03:34, 31 December 2025 (UTC)reply
    Agreed. 1h is reasonable. So9q (talk) 22:32, 1 January 2026 (UTC)reply
  •  Support with the caveat about reasonable time for item enrichment. I find the argument against weak, as empty items are even more of a burden than someone at least trying to add relevant links. (People acting in bad faith is a totally different thing and is a pest regardless if they are creating empty items or adding statements to them.) Ainali (talk) 10:06, 31 December 2025 (UTC)reply
  • As a counter example there were 294 Flower-class corvette (Q404394) produced, collectively they won the Battle of the Atlantic, I have created entries for 290odd of them. The class is widely described, but generally collates the members, with individuals featured in random scattering of articles across Wikipedias and other naval sites, but I'd struggle to find a source that mentions them all. There is a structural need to have articles on the class and all its members, but it would be frustrating if when adding the last few members they were challenged because of poor documentation. If WD is to be comprehensive, it is inevitable it might contain results that were merely entries in lists elsewhere, and so not amenable to formal property assignment. I suppose a line in a table in a URL could be used as a reference or qualifier to a instance of (P31) statement, but it feels clumsy. Vicarage (talk) 10:27, 31 December 2025 (UTC)reply
  •  Support And we should have a uniform way of reminding users of this requirement. --Matěj Suchánek (talk) 11:38, 2 January 2026 (UTC)reply
  •  Strong support This is a very important requirement. I don't see any problems with this as long as we give enough time to add the statements. --Yuriklim (talk) 12:40, 2 January 2026 (UTC)reply
  •  Support Sure, with a reasonable time allowed to get those statements in. ArthurPSmith (talk) 16:50, 3 January 2026 (UTC)reply
  •  Comment Not just statements. Sitelinks are also very important. It is completely acceptable to have an item with nothing but a sitelink. Midleading (talk) 17:13, 7 January 2026 (UTC)reply
  •  Weak support The new item should have at minimum an instance of (P31) or subclass of (P279) with at least one suitable reference that establishes notability. This should be enforced at item creation. Piecesofuk (talk) 10:42, 9 January 2026 (UTC)reply
  •  Support Easy improvement. --Lymantria (talk) 10:17, 12 January 2026 (UTC)reply
  •  Weak support I agree with many of the caveats above. Jerimee (talk) 20:19, 15 January 2026 (UTC)reply
  • Comment: In theory I like the idea, but in practise what would this actually mean?StarTrekker (talk) 09:07, 16 January 2026 (UTC)reply
  •  Comment I very much endorse Mariana Fossatti's comment below: we need flexibility and leeway for underrepresented knowledge, for those communities that haven't had and don't have the privileges and resources to create or maintain sources in the first place. It would also be extremely helpful if it would be made easier and more transparent for beginners how they should do this in practice (i.e. what kind of statements are needed as a minimum; what one typically fills in, etc) Spinster 💬 07:39, 27 January 2026 (UTC)reply

Take-away for policy drafting:

  • What would these statements be? Who would decide? These decisions are usually not made by marginalized communities that often see their knowledge erased. It is completely reasonable to require that the item have enough statements to be recognized and distinguished from another item. And for sure, empty and barely outlined items, impossible for other users to understand and improve, can simply be removed. But introducing notability checkings at the level of required statements can exclude things we don’t want to exclude from Wikidata, and could create a lot of work in order to establish and assess the particular notability criteria for every knowledge domain. Automated statement suggestions, or tools to signal missing statements, are better routes for the improvement of Wikidata. Existing tools, like Recoin, are already helpful for this and could be expanded. Mariana Fossatti (WK?) (talk) 17:09, 23 January 2026 (UTC)reply

Accompanying text: We should give the Wikibase Ecosystem a more prominent place in the text to explain alternatives to adding data to Wikidata outside the notability criteria

[edit]

Votes and comments:

Take-away for policy drafting:

  • The Wikibase ecosystem should not be seen as a repository for the unwanted items. Instead, we should explain its potential to create alternative ontologies and different ways of modeling data for specific needs. It’s great to make Wikibase more prominent and publicize it, not to show to potential editors the Wikidata exit door, but to make them aware about even more exciting possibilities if they want more customization and control over how to collect and organize data communities. Mariana Fossatti (WK?) (talk) 17:10, 23 January 2026 (UTC)reply

Accompanying text: We should encourage people to flesh out and improve existing Items over adding a lot of new Items

[edit]

Votes and comments:

Take-away for policy drafting:

  • Encouraging improvements is always good. However, this path should not make it more difficult to create new items. Both the creation and improvement of items should be equally welcome. Nevertheless, it would be great to make it easier to use tools to improve and maintain items. The Wikipedia home page is a good example of this, as it invites editors to make small improvements. This is an opportunity for new editors to gain experience. Mariana Fossatti (WK?) (talk) 17:12, 23 January 2026 (UTC)reply

Accompanying text: We should make it explicit that people should not create Items about themselves, their business, etc

[edit]

Votes and comments:

Take-away for policy drafting:

It would be very regrettable if this precluded the authors of scholarly works, or their publishers or employing institutes, from creating items about their works. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 13:17, 23 January 2026 (UTC)reply
  • As long as the article has serious references, creating articles about people's identities or activities does not seem to be a major problem. At least, not to the extent that it is a problem on Wikipedia, where editors who write about themselves can introduce serious biases and imbalances into the article's narrative. The structured nature of Wikidata is less vulnerable (not immune) to the introduction of distorted narratives. Mariana Fossatti (WK?) (talk) 17:14, 23 January 2026 (UTC)reply
  • I am late to this discussion, so thanks to Lydia for allowing me a chance to add a comment. I'll disclose that I do communications consulting work on Wikipedia and occasionally on Wikidata, so the matter is of professional interest to me.
My suggestion is that Notability and conflict of interest (COI) are worth considering separately: Notability is about scope; COI is about process. An item either is within scope or it is not, and the identity of who created it should not be a determining factor. Otherwise, we have an arbitrary standard that is exceedingly difficult to enforce. Instead, it's better to focus on requiring disclosure and third-party sources.
I also agree with Mariana that Wikidata is inherently less vulnerable to promotional manipulation: Wikipedia can be subtly promotional through word choice, framing, and emphasis, while Wikidata's structured property-value pairs are more constrained (a statement like Company X → headquarters → City Y is either accurate and sourced or it isn't).
Rather than entangling the two concepts, where COI is concerned I would suggest Wikidata look to how Wikipedia has handled this issue (see: WP:PSCOI) and revisit the RfC about COI from 2015, as well as review the existing policy Wikidata:Disclosure of paid editing and user essay Wikidata:Self-promotion to address this subject with greater care. WWB Too (talk) 21:59, 27 January 2026 (UTC)reply
  • In some Wikimedia projects it is allowed to edit COI content as far as it is acknowledged (=white hat editing). So if they create, say a notable Wikipedia article, why couldn't they add the Wikidata item too? Or add a new item for an old Wikipedia article? The statement was about "creating", not "editing". Was it intentional? I'm sure that if the paid/COI Wikidata editing is forbidden, there will be black hat editors. (I'm a paid editor myself which I've disclosed here). Take a look at how new Wikimedia Commons files are added - what kind of things are asked from the downloader.--Jjanhone (talk) 20:02, 2 February 2026 (UTC)reply
  • I think this wording is extremely vague, I guess my question is not having this rule currently causing an unmanageable issue? As written this could really limit a huge amount of valuable knowledge being added to Wikidata, especially ones that could be considered commercial or private knowledge institutions.
We want this information on Wikidata but explicitly excluding the people who know the most about these collections (and often have the only access to the raw data) doesn't seem like a very good approach at all. Wikipedia and Wikidata are very different, only allowing non involved individuals to add data from some institutions is not a complete solution, especially for datasets which change. Importing whole datasets from knowledge institutions is needed at the very least for completeness of queries. These are a few practical examples I can see that would probably not be allowed under this quite vague wording:
  1. Museums/archives adding items about objects in their collections if they link back to their collections database as a reference.
  2. Universities or other academic institutions adding information about academics who work at their institutions.
I've also seen others suggest that we make Wikidata:Self-promotion policy, I think this not a good idea, especially "Wikidata only includes topics that are already notable in independent, reliable sources", this would create a really significant barrier to underrepresented knowledge where the information is held by only one institution eg minority cultural institutions.
Thanks, John Cummings (talk) 21:50, 17 February 2026 (UTC)reply
It's creating an untenable amount of people creating items about themselves or their companies. This should be distinguished from making an item about something you possess like museum artifacts.
We should not indiscriminately include all knowledge in the world, since we are responsible for the accuracy of our database. Jasper Deng (talk) 21:53, 17 February 2026 (UTC)reply
[edit]

Votes and comments:

Take-away for policy drafting:

Accompanying text: We should encourage people to ask if they are unsure if the Item they want to create is notable, especially if they want to create a lot of new Items.

[edit]

Votes and comments:

  •  Support but where? I would guess a subpage of Wikidata:Notability, like Wikidata:Notability/Questions; please not directly in the Project chat to avoid flooding it even more. --Epìdosis 18:30, 30 December 2025 (UTC)reply
  •  Support but are we prepared to answer them promptly? Unlike Relevanzcheck in dewiki, Relevance Check in WIkidata could be flooded in minutes. I think we need to limit the check to item that related to series or datasets, or generational data. I suggest we make something like basic checklist for notability criteria, yet how do we limit the interpretation? —Yamato Shiya大和 士也 (Talk • Contribs) 12:22, 31 December 2025 (UTC)reply
    Yeah we probably don't want people asking about every single Item. But if they are doing larger sets etc I think it'd be good to make it as easy as we can to get a check on them before creation instead of cleaning up afterwards. Lydia Pintscher (WMDE) (talk) 15:12, 8 January 2026 (UTC)reply
  •  Support Sometimes I wish Wikidata had a centralized “big document” that clearly explains what kinds of data are currently allowed to be entered, specifically what “niches” exist, along with a “model item (P5869)” for each niche. For example: humans, books, stars, species, sports clubs, historical events, wars, accidents, disasters, diseases, anatomical structures, and so on. Ideally, this document would also allow users to revise it by proposing new kinds of data they want to add to Wikidata, perhaps while streamlining the process of proposing the important properties required so that such data can be properly integrated into the Wikidata ecosystem. With this approach, Wikidata would not be built by endlessly adding new properties, but by developing it niche by niche as the smallest “working unit”. For instance, a region is sometimes well known for specializing in producing “something”, whether natural resources or a concentration of a specific industry producing particular goods. People might want to add this connection to Wikidata, such as region X being known for producing apples, while region Y is known for producing movies. Instead of guessing which property to use to add this relationship, they could consult this “big document” to see whether the current Wikidata model already supports such a relationship, and if not, propose it. I imagine this would require an “active duty” maintainer of the document who has a strong understanding of the current Wikidata ontology and can respond to new proposals and questions. Niryhpr (talk) 09:42, 6 January 2026 (UTC)reply
  •  Support A separate subpage for these sorts of requests, also for merge requests and maybe a place to ask whether items should be deleted before they're added to WD:RFD Piecesofuk (talk) 10:58, 9 January 2026 (UTC)reply
  •  Weak support Only if they want to create hundreds of new item. I worry that policies like this, as gentle and reasonable as they sound, can potentially be a real barrier to entry to conscientious new folks. A potential new user might think to herself "ok, I'm not sure, let me think about it some more" and never come back. Jerimee (talk) 20:38, 15 January 2026 (UTC)reply
  •  Support per Piecesofuk. --Wolverène (talk) 08:21, 16 January 2026 (UTC)reply

Take-away for policy drafting:

  • It sounds reasonable to offer better orientation for newcomers to Wikidata and also for more long-term users exploring new knowledge domains. Wikiprojects are a great space for that, and would be good to have access to them from the Notability page. For a single item or just a few, guidelines and resources should be enough. For a bulk upload of new items, it would be desirable to get community insights. Mariana Fossatti (WK?) (talk) 17:16, 23 January 2026 (UTC)reply

Process: We should only let registered users create Items

[edit]

Votes and comments (see also Wikidata talk:Requests for comment/Mass-editing policy#Restrict entity creation to logged-in users):

Take-away for policy drafting:

  • I will not incorporate it into the policy draft. I will do some more research with the developers to see what we can do about throttling the creation of new Items by new accounts while not breaking Item creation from Wikipedia and co. Lydia Pintscher (WMDE) (talk) 11:27, 23 January 2026 (UTC)reply
  •  Oppose Attribution of edits is now managed with temporary accounts across Wikimedia projects, in order to mitigate privacy concerns. Let’s continue with this general privacy-friendly Wikimedia approach, especially to protect those who need a safer edit environment. Mariana Fossatti (WK?) (talk) 17:18, 23 January 2026 (UTC)reply

Process: We should take the intent of the Item creator into account when making decisions about notability

[edit]

Example: Someone gaming the system by creating several Items and connecting them in order to fulfil notability criteria 3, even if the Items would otherwise not be notable

Votes and comments:

  • Establishing intent could become so bureaucratic. Assume good faith for members in good standing. Do we record statistics of how many items each user had has deleted, to trigger different levels of oversight? Vicarage (talk) 16:47, 30 December 2025 (UTC)reply
  • the intent of this is good, but it could also be difficult to apply, as Vicarage says; anyway, probably we should probably try to enlarge a bit "structural need" in order to make it more difficult to game it. --Epìdosis 18:35, 30 December 2025 (UTC)reply
  •  Oppose If someone creates multiple items or redirects the same item to itself, we can already check that and delete/unlink all of the items. Checking their intent seems like a lot of unnecessary work. Ternera (talk) 19:35, 30 December 2025 (UTC)reply
  • This seems to conflate two points.
    • While it is often easy to classify a contribution as promotional, we should avoid going down the path of mind reading and doxxing.
    • We routinely bulk delete groups of items that connect to each other, as not a valid case of N3, but it is harder to investigate. Bovlb (talk) 20:08, 30 December 2025 (UTC)reply
  • I support the idea of this, but am not sure how we implement this in a good way. Perhaps this kid of gaming can be mentioned as an example of bad faith editing? Ainali (talk) 10:34, 31 December 2025 (UTC)reply
  •  Support Like Ainali, I support the idea of this, but am not sure how we might implement it. Jerimee (talk) 20:46, 15 January 2026 (UTC)reply

Take-away for policy drafting:

Process: We should let WikiProjects narrow down (but not expand) the general notability guidelines for their area

[edit]

Votes and comments:

  • Notability so depends on the specifics and data volumes a project is designed for, and seeking agreement could stultify things, so let each project have its own criteria. Vicarage (talk) 16:43, 30 December 2025 (UTC)reply
  • I would specify some minimal criteria for notability criteria established by specific WikiProjects, specifically: they should be easy to find (i.e. at least all linked from one place), easy to read (some kind of standard structure), and they should be established by at least 10 users (to avoid important decisions being taken by too few users). I agree that these thematic guidelines by WikiProjects should be only restrictive in comparison with the general ones. --Epìdosis 18:39, 30 December 2025 (UTC)reply
  •  Support I think this is a great idea, but I think we need to figure out a way to be able to point from the general criteria to WikiProjects that have found consensus for narrower ones in their field. (We don't want to spring this as a surprise on someone who went to the general page and read it.) Ainali (talk) 10:45, 31 December 2025 (UTC)reply
  • We should be wary of letting one wikiproject declare that something is not notable, when another wikiproject considers that it is. Conversely, we should be wary of a wikiproject existing just to declare that everything in a given field (Wikiproject SEO Experts? WikiProject YouTube Influencers?) is notable. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 19:02, 31 December 2025 (UTC)reply
    • The point of multiple WikiProject viewpoints are a good one. And I think that it can be solved by not allowing a WikiProject restrict another one. That is, if one WikiProject finds some type of concept notable, but not another, the second WikiProject can abstain from creating items, but not stop the first one. Ainali (talk) 20:01, 31 December 2025 (UTC)reply
    • Regarding the point of declaring everything notable, they still shouldn't be able to expand the general notability policy. Ainali (talk) 20:01, 31 December 2025 (UTC)reply
  • I have always liked the ideas of WikiProjects, or in general, associations by topics (music, Sweden, influencers, ...) as place to gather all sorts of resources (showcase items, list of properties, do's and dont's, ...). But now I am a little reserved because this could introduce sort of a backdoor to the general policy (i.e., a possibility to overrule what we are trying to reform right now). --Matěj Suchánek (talk) 12:20, 2 January 2026 (UTC)reply
    Do you even have this worry if we only allow them to narrow the general criteria? Do you have an example of something that could be problematic? Lydia Pintscher (WMDE) (talk) 15:18, 8 January 2026 (UTC)reply
    For example, in far-off foreign lands over there (OpenStreetMap), there are several cases where local-specific groups create and enforce their own local rules that actually contradict the global rules agreed upon earlier. Normal unaffiliated users might get confused by these conflicting rules, but in most cases, local groups enforce them more strictly than the “global police” do, and they can even band together to push back against site-wide moderators. Rtnf (talk) 06:52, 9 January 2026 (UTC)reply
    It can still be source of disagreement between admins and regular community members with a field of interest ("we consider this notable regardless"). I'd prefer if we had rules that the whole community agreed upon. Scoped guidelines could be handled by appendices/addenda. --Matěj Suchánek (talk) 10:05, 17 January 2026 (UTC)reply
  •  Comment I don't see the need for notability policy expansion. The two benefits listed at The benefits and problems... are truly valuable benefits. Problems should be addressed without sacrificing ease of use. Jerimee (talk) 21:02, 15 January 2026 (UTC)reply
  •  Oppose. Sounds like an attempt to implement something for which I would criticize Wikipedia the most -- like the decisions on what is notable made by some certain group of people (who are BTW not experts in their area IRL and not obligated to be), can be more valuable than the general notability policy. No, it is wrong. --Wolverène (talk) 08:56, 16 January 2026 (UTC)reply
  •  Oppose Inclined to agree with Wolverène here.StarTrekker (talk) 09:06, 16 January 2026 (UTC)reply
  •  Question From what I've gathered from my work here at Wikidata, there are already subject areas with modified notability criteria. As far as I know, schools are considered notable per se. Such per se notability would no longer be possible under these new rules, would it? --Gymnicus (talk) 10:57, 16 January 2026 (UTC)reply
  •  Oppose per Wolverène. --Prototyperspective (talk) 18:06, 27 January 2026 (UTC)reply
  •  Oppose This would compound systemic bias against minoritised knowledge. — OwenBlacker (talk; please {{ping}} me in replies) 13:55, 2 February 2026 (UTC)reply
  •  Support (Sorry for being so late in responding.) I generally support this idea. The notability guidelines are quite high-level so there is benefit in making them more specific in general areas. If a project makes their guidelines such that some other aspect of Wikidata is negatively affected, then they would be subject to discussion and revision. Peter F. Patel-Schneider (talk) 15:06, 24 March 2026 (UTC)reply

Take-away for policy drafting:

Round 1: Other questions and discussion

[edit]
Questions and discussion that was used to draft the new policy

How can we rely more on tooling and automation for notability checking and processing?

[edit]

Do you have additional suggestions for how to tighten/clarify criteria 2?

[edit]
  • Copied from above so that they aren't lost: https://www.wikidata.org/w/index.php?title=Wikidata:Project_chat&oldid=2449188556#%22Serious_and_publicly_available_sources%22_and_self-promotion and https://www.wikidata.org/w/index.php?title=Wikidata:Notability/sandbox&oldid=1312939910 should be considered as starting points. --Epìdosis 18:40, 30 December 2025 (UTC)reply
  • I'm a fan of the suggested policy update laid out by @Bovlb in the conversation linked above. Ternera (talk) 19:37, 30 December 2025 (UTC)reply
    Me too So9q (talk) 22:09, 1 January 2026 (UTC)reply
  • Extreme case: remove criteria 2 altogether and just use 1 and 3: criteria 1, an item must have a Wikipedia article; criteria 3 an item "fulfills a structural need" for example, an item is a named entity contained within a Wikipedia article. Alternately criteria 2: an item passes criteria 2 if it is the subject of a reputable encyclopaedic or biographical article, book etc., or is the subject of a widely reported event. A reputable source must have a Wikidata item that passes criteria 1 or 2. Also, non-alias Wikipedia redirects should also pass criteria 2 Piecesofuk (talk) 11:03, 9 January 2026 (UTC)reply
  • For companies, we should clarify how to treat sources that are serious, but indiscriminately include every company in the country, such as government registries. There's currently a debate on Admininstator Noticeboard about it: [1]. I think the intent is that they should not establish notability?
  • For shcolarly articles, we also need to clarify what's sufficient to establish notability. Currently there are bots mass importing article items with just DOI, does this mean that DOI is enough to establish notability? What if there's no DOI but equivalent other identifier, such as Russian E-library ID? MSDN.WhiteKnight (talk) 15:47, 10 January 2026 (UTC)reply
    A DOI alone must not be sufficient to establish notability, given that some organizations like Zenodo indiscriminately assign a DOI to any document uploaded to their service. Omphalographer (talk) 21:53, 18 February 2026 (UTC)reply
  • Maybe the general issue that motivates the proposal of tightening the notability criteria (the problems identified by Lydia) could be addressed at the level of data modeling and in Wikiprojects scope, where we can find more concrete definitions and boundaries for specific knowledge domains. For instance, WikiProject Books use a “a two-layer framework, consisting of work and edition”, keeping outside of their scope particular volumes. It is clear that Wikidata can’t be a book inventory, even when library catalogues have IDs for each volume they host. Maybe for this delimitation problem, it is useful to reflect again about Wikidata scope and goals, having in mind the simple idea that Wikidata is not an exhaustive inventory for each knowledge domain. And perhaps we should take another look at the idea of a general knowledge database, considering it afresh with the benefit of years of experience. It should be one that doesn't leave behind marginalised knowledge and doesn't reproduce unjust biases, while keeping a broad but also simple and transparent delimitation criteria. Mariana Fossatti (WK?) (talk) 17:25, 23 January 2026 (UTC)reply
    With due respect, a blanket reform is still needed. The community will not be able to track every decentralized tightening of the notability guidelines and the project-wide one is still much too lenient. Jasper Deng (talk) 20:52, 17 February 2026 (UTC)reply

Is there anything we missed?

[edit]
  • When we will discuss the draft of the new notability policy, in the same discussion we will also necessarily have to discuss about the items already existing that will be outside the revised notability criteria, and specifically: 1) how to find them; 2) what to do with them (keep them for some reason; simply delete them; delete them and move them elsewhere [where specifically?]). --Epìdosis 18:42, 30 December 2025 (UTC)reply
  • Maybe we should consider somewhere the case of items which met criterium 2 at the time of their creation and does not meet it anymore (i.e. they are based on a reliable online source which has perished in the meanwhile and has not been archived, or only partially); do we want to keep them? Probably not. Cf. Wikidata:External identifiers/Obsolescence (draft) for more context. --Epìdosis 18:54, 30 December 2025 (UTC)reply
  • The Wikibase Ecosystem has matured, with Wikibase Cloud and Wikibase Suite now being viable alternatives for hosting knowledge graphs that are tightly interconnected with Wikidata
    Is this based on the needs of the Wikidata community or is it a pipe dream? From what I have understood talking to the Scholia team and knowledge experts this idea about ontologic federation is currently an unknown thing.
    The Scholia team decided to ABANDON WDQS after the graph split of scholarly items because the federation in SPARQL did not work (timeouts, hard to know where the items reside, etc.). I consider the Scholia team to be SPARQL experts and if they can't get it to work I very much doubt anyone else will. Wikibase Suite in it's current state is NOT able to reliably integrate with Wikidata in my opinion. I tried setting up a Wikibase.cloud wiki multiple times and get it integrated, but the federation UX is nowhere to be found.
    My current conclusion is thus: the "ecosystem of connected Wikibases" is not a feasible idea that has gained any mentionable traction. Even if the federated properties + values were to gain adoption it would result in a split of the community and Wikibase Suite does not support RDF streaming so the triples cannot be integrated e.g. by combining multiple streams in a single QLever instance.--So9q (talk) 22:08, 1 January 2026 (UTC)reply
    • This could all work, but only if there was a good way of really federating subsidiary Wikibases with Wikidata, and an easy way to get changes out of the subsidiary Wikibases, and some way to set up SPARQL services for the combination of Wikidata and one or more subsidiary Wikibases. But I don't see much, if any, action on any of these. Peter F. Patel-Schneider (talk) 19:10, 2 March 2026 (UTC)reply
  • I don't see scaling Wikibase anywhere above. I recently started to work on wikibase-backend (built using Python, FastAPI) which scales to 1bn+ items and it is as of writing nearing MVP state and already capable of:
    • CRUD entity operations
    • scaling the metadata database to 1bn+ entities (using sharding in Vitess)
    • scaling to 10bn+ revisions (S3 compatible backend based on Ceph)
    • scaling to 1 trillion statements (S3, deduplicated)
    • entity locking and archiving is supported
    • entity redirects are supported
    • full RDF output is supported (98% complete as of today 🥳)
I added cost estimates to the repo and they are looking very promising. The current monolithic Wikidata architecture cannot scale to 10x the current number of revisions. It has reached end of life a number of years ago and neither WMDE nor WMF has done anything about fixing the root cause (band-aids, investigations and disaster plans don't count 😉).
This new backend if finished and implemented may very well be a game-changer for the future of Wikidata and notably not one that break current tools like the "ontological federation" effectively would. It most probably also be easier for the new Wikidata Platform team to operate in a sustainable way compared to the current legacy architecture where manual pooling and installation of servers is needed.--So9q (talk) 22:08, 1 January 2026 (UTC)reply
One pattern is to upload photos to Commons, create a category there, and then use the existence of that category to justify a Wikidata item. In some cases, the existence of the Wikidata item is in turn cited on Commons as a reason not to delete the category, creating circular notability between projects.
Another pattern is to create articles (often machine-translated) on smaller Wikipedias. These projects often have limited administrative capacity, and cleanup can take a long time, during which the article’s existence is used to satisfy N1.
These approaches are inexpensive for the creator but time-consuming for communities to resolve.
I’m not sure what the best long-term fix is, but one small improvement might be to change N1.4 by dropping the first word, so that it applies to all items rather than only category items.Bovlb (talk) 16:17, 4 January 2026 (UTC)reply
Can't items with a commons category can still be deleted under N1:4? Also, spam articles from small projects can be added to meta:GSR for global sysops to delete if there are no active admins. The request page is closely monitored and spam is deleted within hours usually. Ternera (talk) 03:38, 6 January 2026 (UTC)reply
No. Only category items can be deleted under N1.4.
Not all smaller projects are handled by global sysops. There's a middle group of projects that choose to be independent, but unfortunately struggle to handle backlogs. Bovlb (talk) 05:21, 6 January 2026 (UTC)reply
  • I see that the reuse criterion does not seem to go through. I do not know myself whether this is good or bad, but probably the most known / the biggest case of reuse is Open Street Map which uses Wikidata to translate all features it shows into many languages (which is important because it uses the native language for every country - I have just been one week in Cairo, and it is difficult to use without fluent command of the Arabic alphabet). We probably should have a position on which entities are important for us and which ones we want to host. Note that the vast majority are not related to self-promotion, though some (like shops) can be.--Ymblanter (talk) 20:24, 5 January 2026 (UTC)reply
  • I didn't see any mention of unpatrolled edits. There are now so many unpatrolled new items that even blatant spam or vandalism can survive a month and then just sit there for years until someone stumbles across it. My preference would be for these items to remain tagged as unpatrolled permanently until someone actually checks them. —Xezbeth (talk) 09:48, 6 January 2026 (UTC)reply
    I would  Support this. Epìdosis 11:23, 6 January 2026 (UTC)reply
    This is a technical limitation that cannot be lifted. But there are bots taking snapshots, like User:Pasleim/notability. --Matěj Suchánek (talk) 19:01, 7 January 2026 (UTC)reply
  • How the Wikidata community should assess, assure, and support the notability of editing activities that are part of projects funded by the Wikimedia Foundation. When there is less clear explanation or visibility of the project context, such editing activities may be misinterpreted as spam or non-notable, despite being carried out within an approved project. Example of such happenings are here and here. We should be able to "guide" or "advise" those projects, like how to filter data that are notable enough, or to help them create proper attribute. Usually what happens are only request for deletions or questions of notability in user talks. I think we should enforce that such projects need to have clear plan on the method they used, the data, the source, and perhaps example plan of final result that they want to achieve.Happy New Year 2026!Yamato Shiya大和 士也 (Talk • Contribs) 09:08, 8 January 2026 (UTC)reply
  • How Wikidata should allow sufficient wiggle room for items originating from countries or regions with limited access to standardized reference infrastructures, such as Indonesia or other Global South countries. In the absence of widely used authority identifiers or centralized reference systems, for example VIAF, ISNI, national libraries, or the Library of Congress, data that are locally valid in context may still fail to meet Wikidata’s notability criteria, which can affect items such as Indonesian landmarks, for example, religious structures.Happy New Year 2026!Yamato Shiya大和 士也 (Talk • Contribs) 09:08, 8 January 2026 (UTC)reply
  • In this and the mass-editing RfC, and the underlying move towards federated Wikibases, I am extremely concerned about knowledge equity, and about the sustainability of underrepresented knowledge. We risk to even further replicate existing biases in knowledge production, to continue to keep communities with less or no resources out of the picture, and to deprioritize knowledge and data that has been left out of the historical record due to structures of power and privilege. How many Wikibases with underrepresented knowledge will still be up and active, will still have resources and critical mass in, say, 10-15 years? Spinster 💬 10:31, 8 January 2026 (UTC)reply
  • Barriers to contribution for those who actually have valuable stuff to contribute. In my work (on cultural subjects, even in well-resourced countries), I see all the time that Wikidata is seen as a difficult and high-barrier project to contribute to. The current RfCs risk to result in even more barriers. It would be helpful to focus: create very strict and punitive barriers to those who actually actively undermine Wikimedia goals; and on the other hand aim to become much more welcoming to more contributors who have Wikimedia-aligned knowledge to contribute. Spinster 💬 10:31, 8 January 2026 (UTC)reply
  • With the increase of Wikidata items create for solely SEO purposes, we should create an essay to explain Wikidata policies regarding this so admin could just refer to them when deleting items. Happy New Year 2026!Yamato Shiya大和 士也 (Talk • Contribs) 15:24, 8 January 2026 (UTC)reply
  • I think this needs to be considered alongside criteria 1 and 3 and also what "serious and publicly available references" means. Also how do changes affect external projects: https://www.wikidata.org/wiki/Wikidata:For_developers/Showcase Integration with Open Street Map https://wiki.openstreetmap.org/wiki/Wikidata and small academic projects like https://www.wikidata.org/wiki/Wikidata:WikiProject_Scotland%27s_Accused_Witches and on focus list of Wikimedia project (P5008) There should also be clarity at other sources of item creation, for example Open Refine, QuickStatements, https://meta.wikimedia.org/wiki/WikiShootMe#Create_new_items and https://mix-n-match.toolforge.org/#/ Piecesofuk (talk) 11:13, 9 January 2026 (UTC)reply
  • I think there is a larger impact on GLAM work that I can't see yet in the discussion above: 1) my perspective is that I used to work at a museum in the UK where the collection was catalogued, but there was no online collection and it was very little published. With closer constraints on notability objects it sounds like parts of this collection couldn't be integrated. I imagine this would be an issue for others outside of the UK. 2) I have another concern, which relates to the short-term state of GLAM funding in the UK - say an organisation wanted to do some Wikidata work, and got a 1 year grant to do so, and set up a Wikibase, in my experience once the 1 year post holder is gone it's unlikely that anyone in the organisation would either have the skills or interest to continue. I would strongly question whether this policy change would be sustainable for collaborators. Lajmmoore (talk) 14:19, 27 January 2026 (UTC)reply
    & a general point, what happened to wanting to represent the sum of all human knowledge? It feels like this is a contraction of purpose Lajmmoore (talk) 14:21, 27 January 2026 (UTC)reply

Round 2: Context

[edit]
Context for initial policy proposal

Below is the proposal for the new policy based on the discussion in round 1. Please add your votes, questions and comments below.

There are office hours on the following days to talk more about the policy if you are interested (though the binding votes and discussion need to happen in this RfC):

The calls will happen at https://meet.jit.si/WikidataNotability

Round 2: Policy proposal

[edit]
The initial policy proposal text

Wikidata is a knowledge graph, providing general purpose data to the Wikimedia projects and the world at large. Wikidata’s data is intended to provide a basic understanding of the world for humans and machines.

This policy defines which entities can be described in Wikidata.

An item is acceptable if and only if it meets at least one of the three criteria below:

  1. It contains at least one valid sitelink to a page on Wikipedia, Wikivoyage, Wikisource, Wikiquote, Wikinews, Wikibooks, Wikidata, Wikispecies, Wikiversity, or Wikimedia Commons.
    1. <subpoints of criterium 1 remain unchanged>
  2. It refers to an instance of a clearly identified conceptual or material entity that is described using serious and publicly available references. The minimum criteria:
    1. At least one external identifier: The subject is associated with one or more identifiers that are assigned or curated by an independent authority, and that meaningfully distinguish the subject from others. Identifiers that are self-assigned, or can be obtained without editorial review (such as typical social-media profiles) are not sufficient on their own. OR:
    2. At least one reference, independent from the subject: The subject is the topic of significant, independent coverage in serious and publicly available sources that support a meaningful identification and distinction from others. Routine listings, directory entries or being quoted in an interview about another topic do not constitute significant coverage on their own, and additional independent sources might be required. Self-authored descriptions, and marketing or promotional material are not considered serious and independent sources.
  3. It fulfills a structural need, for example: it is needed to make statements made in other items more useful.

New items must have the statements necessary to be clearly identified or sitelinks that show that they meet at least one of the above criteria within 24 hours.

Creating items about yourself, your close relatives, your business and other entities you have a conflict of interest with is not permitted.

Please consider

[edit]

Wikidata’s data is maintained by a community of people like you. Please be considerate of their time and effort. This is important for the long-term health of the project. Before creating a lot of new Items:

  • Spend some time editing existing Items to get an understanding of data modelling and best practices first. We value quality over quantity.
  • Think about where and how your data will be used. If your data is used in a Wikimedia project or another application, chances are higher that it will be kept in good shape long-term and Wikidata’s overall data quality stays high.
  • Consider what the complete set of Items would be that you are working on. Wikidata is not the right place to describe all stars in the universe, every human on earth, etc. Narrow it down to a reasonable set of notable entities and discuss with the community in case of any doubt.
  • If you intend to create a lot of new Items and are not absolutely sure if they meet the notability criteria, please ask for additional input at Wikidata:Notability/Questions.
  • Do not try to game the system. Such efforts are frowned upon and will only lead to your Items being deleted.

Knowledge equity

[edit]

We do acknowledge that certain knowledge has been and is being structurally marginalized. This leads to less coverage in reliable sources and therefore an increased barrier to demonstrating notability. In our effort to make more knowledge accessible to everyone, we take this limitation into consideration when making decisions about the notability of an entity.

Options besides Wikidata

[edit]

If your data does not fit Wikidata’s notability criteria, we want to help you find the perfect home for your contributions. Wikibase Cloud and Wikibase Suite might be a better fit for your data. Together with Wikidata, these form the Wikibase Ecosystem, a network of specialized communities, each collecting and maintaining data for the world in their area of focus.

Data in any Wikibase instance becomes part of the Linked Open Data Web and accessible worldwide. In addition you have more flexibility in how to model your data and who gets to edit it. Visit wikiba.se to explore which platform best suits your needs, and check out our showcase to see examples of what other communities have built.

Other entity types

[edit]

This notability policy covers Items. The notability criteria for Lexemes can be found at Wikidata:Lexicographical data/Notability.

Round 2: Votes, questions and discussion

[edit]
Discussion about the initial policy proposal
  • @Lydia Pintscher (WMDE): The self-promotion page is currently an essay. Can we formally elevate it to a policy too? Let's be explicit about proposing that.--Jasper Deng (talk) 10:25, 19 February 2026 (UTC)reply
    @Jasper Deng Good point. Would that be a separate RfC? Lydia Pintscher (WMDE) (talk) 10:26, 19 February 2026 (UTC)reply
    @Lydia Pintscher (WMDE): For time's sake, I think it's okay to be in this RfC as well, as long as it's under its own section and proposal. There also must be a provision for what to do with the notability policy if self-promotion is not enacted as a policy.--Jasper Deng (talk) 10:27, 19 February 2026 (UTC)reply
    I've added a section for it now. Lydia Pintscher (WMDE) (talk) 15:38, 19 February 2026 (UTC)reply
  • I like "identifiers that are assigned or curated by an independent authority" but I also wonder what that means in practice. Is a museum an independent authority for art they own? Would it make sense to mention that the identifier should at least have Wikidata property for an identifier that suggests notability (Q62589316) (and never only Wikidata property for an identifier that does not imply notability (Q62589320))? Ainali (talk) 15:26, 19 February 2026 (UTC)reply
    I think that this is very problematic criterion, because PR agencies will know that they need to by an ISNI before getting a QID. In many countries (perhaps more than not) issuing free PIDs by GLAM institutions is a very politicized process and excludes minorities, politically subversive content and their creators. I do not find this criterion good. Adaniel (talk) 08:30, 1 March 2026 (UTC)reply
    I think what we want is that people can't just get the ID themselves for themselves. Reading over it again this is maybe covered sufficiently with the rest of the section? We could remove "that are assigned or curated by an independent authority". What do people think? Lydia Pintscher (WMDE) (talk) 11:24, 5 March 2026 (UTC)reply
  •  Support I like the new proposal as it is, I think it reflects well the consensus from the earlier discussion. ArthurPSmith (talk) 17:08, 19 February 2026 (UTC)reply
  •  Comment I agree with ArthurPSmith, but with one fundamental doubt (related to the question raised by Ainali): "identifiers that are assigned or curated by an independent authority" should be clarified, probably with a reference to the presence of instance of (P31)Wikidata property for an identifier that suggests notability (Q62589316) and the absence of instance of (P31)Wikidata property for an identifier that does not imply notability (Q62589320), and this is probably uncontroversial ... but what will probably be problematic will be assigning one of these two to the properties not already having them, and whilst this is probably easy for most of them, I am afraid for the IDs regarding bibliographic entities: the bibliographic data of a book/article described in a library catalogue and/or bibliographic database are inside this notability policy, if they have no incoming links from other items (i.e. criterium 3)? As of now DOI (P356), ISBN-10 (P957) and ISBN-13 (P212), as well as others like e.g. PubMed publication ID (P698) or DNB edition ID (P1292), and in general most of the "bibliographic IDs" (I will try to make precise statistics about this) don't have neither, so the situation of items relying solely on them (I guess some 40 M) remains ambiguous, which is surely not ideal, since we are exactly trying to make it clear the point. As we have opened a thread below regarding the status of WD:Self-promotion, I think opening another subparagraph for the discussion regarding the notability-status of bibliographic IDs would be highly recommendable (or a separated RfC). I also add a related question (although it impacts much much fewer items): if I have a printed bibliography listing some hundreds editions of notable works (e.g. tragedies of Sophocles), or hundreds of books/articles regarding a certain theme, and I want to create all their items using this printed authoritative source, if I understand correctly the policy would allow it; I tend to think this is a good thing, but I would like to make sure this is also the general consensus. --Epìdosis 18:18, 19 February 2026 (UTC) BTW, it should become mandatory ASAP to assign to all newly-created external-ID properties instance of (P31)Wikidata property for an identifier that suggests notability (Q62589316) or instance of (P31)Wikidata property for an identifier that does not imply notability (Q62589320)reply
    +1 on the last sentence; we should even make this mandatory in the property proposal and not mark them as ready unless it has been filled in. Ainali (talk) 18:44, 19 February 2026 (UTC)reply
    @Epìdosis I would be usefull to add a page like en:Wikipedia:Reliable sources/Perennial sources, in case there is contestation one way or another. Like I seen a RfD using ‎Organ Index ID (P13889) today and I was not sure if it establish or not the notability. Fralambert (talk) 20:59, 19 February 2026 (UTC)reply
    There are identfiers such as Wikimedia username (P4174) that clearly don't suggest notability, others that I think should be accepted as notable (NRHP reference number (P649) and similar) then those that are less certain - TOID (P3120) I think is notable but should generally be added to existing items, or new items with other identifiers, GSS code (2011) (P836) notability probably depends on prefix, GNS Unique Feature ID (P2326) I think is generally notable but can be deleted if we are unable to identify what the ID refers to. Then there are DoBIH Number (P6515) (does not suggest notability because it's self-published or user-generated, but in almost all cases is notable), GeoNames ID (P1566) (does not specify, user-generated so not an indication of notability by itself but ID and history may point to other sources). If DOI (P356) can suggest notability, it can only be for certain publishers, journals or prefixes, or with other identifiers, because it can be used for a preprint (Q580922), predatory publishing (Q29959533), or user-generated content (Q579716) (Wikidata:Requests for deletions/Archive/2025/01/16#Q58746840) - PubMed publication ID (P698) seems to be more selective. There seems to be no consistency with library IDs: Library of Congress Authorities ID (P244) says it suggests notability but J9U entity ID (P8189) says it does not but the reason for that is unclear. Peter James (talk) 22:56, 19 February 2026 (UTC)reply
    There are 87 Wikidata properties that have, if my understanding of the properties is correct, contradictory statements on notability: P2429:eventually complete (Q21873974) and P31:Wikidata property for an identifier that does not imply notability (Q62589320) (https://w.wiki/HuoE). Peter James (talk) 23:48, 19 February 2026 (UTC)reply
    Good find! That almost look like a to-do list. Ainali (talk) 06:57, 20 February 2026 (UTC)reply
    Worth to mention at this point that the pair of items Wikidata property for an identifier that suggests notability (Q62589316) and Wikidata property for an identifier that does not imply notability (Q62589320) is a community effort without coverage by any policy at this point. Anyone can add it to or remove if from a property page, and chances are that nobody would object. This definitely needs a clean-up, and then some continuous policing effort to keep it usable. —MisterSynergy (talk) 14:26, 20 February 2026 (UTC)reply
    Sure, I very much agree on this! Epìdosis 15:15, 20 February 2026 (UTC)reply
    I will add a new section further down to discuss this. Lydia Pintscher (WMDE) (talk) 10:18, 22 March 2026 (UTC)reply
  • On the "#Knowledge equity" section: as much as I am aware, this has barely been discussed here yet. While I can generally acknowledge the situation and support that approach, I am not aware how to take this limitation into consideration when making decisions about the notability of an entity. Do we have any guidelines how to handle these cases, from admin perspective? —MisterSynergy (talk) 21:27, 19 February 2026 (UTC)reply
    @MisterSynergy There are many resources about knowledge equity, and there is a lot of experience—and frustration—across Wikimedia projects regarding support for editors working on marginalized topics. However, this has not yet been consolidated into guidelines for administrators, patrols, and for the daily editing/checking work on Wikidata. I see this as a good opportunity to start with something useful and pragmatic in this line. I can to volunteer to outline some ideas to contribute to the conversation. Mariana Fossatti (WK?) (talk) 14:07, 20 February 2026 (UTC)reply
    Can we please have a robust and actionable policy here before we vote this section into policy status? I am otherwise afraid that we create a huge loophole that would allow many nefarious actors to place their undesired content here based on marginalization claims, while actually marginalized knowledge does not have sufficient support for reliable integration at Wikidata based on the proposal above alone. —MisterSynergy (talk) 14:31, 20 February 2026 (UTC)reply
    Hi @MisterSynergy I do support that we should have a robust policy on knowledge equity, and I commit to work in that direction (and if you want to join me, and others, that would be fabulous). But I don't think we need to have it done to approve the notability policy reform. While I share the general concern about promotional content and spam, I have rarely seen marginalized knowledge being used as an excuse to introduce it (all those tech entrepreneurs creating items for their new AI startup that you can see in Recent Changes all the time, don't seem to be claiming that we should accept them because they belong to marginalized communities). The proposed text, given its place in the policy, does not seem to be a checkbox that automatically grants notability, but rather an aspect to be considered with caution and common sense. Wiki editors have a long collective experience of applying common sense and discussing civically until we reach a consensus on notability. I believe that we can continue to be guided by these principles while we develop a more robust guideline on marginalized knowledge and knowledge equity. Having at least one mention of it in this policy is a big step we should support. Mariana Fossatti (WK?) (talk) 18:50, 23 February 2026 (UTC)reply
    Agreed, marginalization is currently not a common claim for those who seek entry but are denied/undesired here. However, in past years there was a clear trend to comply with the current notability policy in order to place promotional content here, as promotional editors understood how they could game the system. We are basically here discussing an RfC because the current form of the policy really needs to be overhauled and clarified based on that experience.
    As for what is sufficient regarding knowledge equity: I am serving as admin here at Wikidata for almost a decade, and my deletion volume is usually pretty high (I have logged more deletion events than anyone else, by quite some margin). I currently do not feel prepared to consider knowledge equity in deletion discussions and I would welcome a comprehensive guide how to handle such cases. The (proposed) policy requires me to do level up my skillset here. —MisterSynergy (talk) 20:24, 23 February 2026 (UTC)reply
    I am going to add a section further down where we can collect ideas for how to address this. Lydia Pintscher (WMDE) (talk) 09:48, 22 March 2026 (UTC)reply
  • For book editions, newspaper or periodical issues etc. hosted on Wikisource, sitelink for main namespaces might not always be present. The Wikisource workflow requires transcription, proofreading and transclusion of the entire file to generate a Wikisource sitelink of the main content and that requires time. Presence of Wikisource index page URL (P1957) with or without Wikisource sitelink should be sufficient to establish its notability. Some of the Wikisource related tools are also based on the presence of this property like Sangkalak (Q123617697), Wikisource Reader app (Q133540085) etc. So, IMHO, this should be documented in the 1st criteria of notability as a footnote. -- Bodhisattwa (talk) 13:37, 20 February 2026 (UTC)reply
    Are there any objections to this? Lydia Pintscher (WMDE) (talk) 11:37, 5 March 2026 (UTC)reply
    I agree with @Bodhisattwa, because creating a main namespace on Wikisource need more time, it would be better if property:P1957 could be included in the first criteria. And index on Wikisource using edition/book items to connect to files in Commons or other Wikisources (if translated). Empat Tilda (talk) 18:13, 7 March 2026 (UTC)reply
  • Finding external identifiers for organisations can problematic. They often don't naturally have a master organisation that mentions them, and if they do, that's often declaration by the organisation itself or a mention on the other organisation's website, but not a formal identifier. For example of the two hundred-odd caving clubs in Britain, perhaps 20 are notable because of their heritage and detail of their research publications, and have content we want to link to with described by source (P1343). Their parent body, British Caving Association (Q4969613) has a page listing them, but there's no formal identifier, and I guess member of (P463) is not acceptable. Only if they are also charities they might have proper identifiers. The same applies for companies and their trade bodies, where paid general directories are of dubious worth. Vicarage (talk) 13:54, 21 February 2026 (UTC)reply
  • I strongly object to the first point of the notability criteria stating that a link to Wikimedia Commons is sufficient for notarization. We had lengthy discussions before we succeeded in ensuring that administrators do not automatically consider Wikimedia Commons to be notable, and I don't want to lose that. Wikimedia Commons lacks any real relevance criteria. --Gymnicus (talk) 23:18, 23 February 2026 (UTC)reply
    The wording in the draft is exactly the wording that is in the current policy. Commons sitelinks are restricted in the way I believe you are asking for in the subpoints of criteria 1. I just omitted them in the draft because they are untouched. Lydia Pintscher (WMDE) (talk) 11:41, 5 March 2026 (UTC)reply
    @Lydia Pintscher (WMDE): It's true that officially, nothing has changed regarding the notability rules since 2013. However, there have been certain—let's call them—guiding decisions from the administrators, and I recall that administrators didn't always classify a Wikimedia Commons link as requiring notability. But administrators would have to confirm whether my recollection is correct. --Gymnicus (talk) 22:59, 5 March 2026 (UTC)reply
  • Every item that enters Wikidata can be roughly classified into a category, such as humans, organizations, businesses, species, stars, chemical substances, books, songs, movies, and so on. I think we can enumerate all the categories that are currently 'allowed' in Wikidata and devise specific notability criteria for each. For each category, we can provide real-world examples of what is notable and what is not. I think we could start a WikiProject to build this kind of 'master guideline,' ideally even reviewing each item from Q1 to Q????. Rtnf (talk) 03:32, 24 February 2026 (UTC)reply
  •  Oppose Aside from the huge implications this policy will have for underrepresented and marginalized knowledge, I'm concerned about the process in itself, particularly where the outcomes are so unclear. We all know that once a policy is settled and enforced, it's extremely hard to roll it back or revert it; it becomes the new standard regardless of its flaws. As an example, the research made by Art+Feminism that spoke to the problems of some of the policies in English Wikipedia has been extremely hard to implement.
How decisions are made is almost as important as the decisions that get made. The way in which this discussion is unfolding makes it very hard to achieve a fair & equitable outcome. The current format, conducted entirely in long-form English, is a massive barrier for a project as multilingual as Wikidata. It effectively silences a huge base of active contributors who have valuable opinions but can't or won't participate in this specific format or language (or that might not even be aware that this is being discussed right now, because they're mainly focused in contributing on the project). Right now, we're looking at a situation where maybe 10 people are deciding the future for thousands of users, including newcomers, that we need more than ever.
If this policy ends up punishing marginalized or otherwise relevant knowledge further down the line, it won't matter that it was "discussed" if the discussion wasn't actually accessible to the people it affects most. Is this actually the best way to approve project policies in 2026, especially one that will have such a tremendous impact on how people can contribute?
If, nevertheless, such a small group moves forward with this, I suggest safeguards are built to prevent long-term damage. First, the policy should only be effective for a set trial period, after which it must undergo a review. Annual reviews on the policy should be conducted to make sure that it remains fit for purpose. We need an immediate "emergency" review trigger if we see the policy disproportionately affects specific groups, like non-native English speakers or newcomers. We also need a clear process to track the "before and after" of this implementation. Without hard evidence of how this affects and changes participation, we won't have the tools to fix it further down the line when the harm becomes clear. Finally, the door needs to remain open for outside perspectives. If an affiliate or an external research group produces solid, evidence-based research on the impacts of this policy, that data should be officially taken into consideration in future discussions about the policy. Scann (talk) 14:36, 24 February 2026 (UTC)reply
In terms of safeguards, I wouldn't discard the possibility also of running a Central Notice Banner and conduct a proper voting process, similar to how stewards and other bureaucratic processes are run. A policy with such an implication can't be decided between a small group in a page that no one actively monitors; it should be publicly communicated and run as any other important bureaucratic process. Scann (talk) 14:56, 24 February 2026 (UTC)reply
I totally agree. So9q (talk) 00:38, 3 March 2026 (UTC)reply
Supporting Scann's comments here, as the existing structural marginalisation of many groups' knowledge would be further baked in by this process Avocadobabygirl (talk) 20:46, 24 February 2026 (UTC)reply
I strongly agree with all the points raised by @Scann! Particularly their demand for an emergency review trigger to alert communities very likely not aware of discussions and policy changes that will have potentially huge impacts on their continued presence and representation on Wikidata (and beyond). OpenArabicPE (talk) 10:18, 10 March 2026 (UTC)reply
I think that the question is whether this is an improvement and good enough. I say yes to both. The policy should be open to update and reform, which would be the way to fix any problems that it has. Peter F. Patel-Schneider (talk) 15:10, 24 March 2026 (UTC)reply
  • Creation and improvement of Wikidata items is often an important step towards identifying what's notable. In training new editors I've always emphasised that this work is collaborative, and the bit of information they add might be supplemented by someone else until there are suddenly enough details and sources for a Wikipedia article to be created. Preventing the creation of items (especially when the criteria reinforce structural issues regarding marginalised knowledge) will make this iterative improvement harder to kick off. Avocadobabygirl (talk) 21:22, 24 February 2026 (UTC)reply
    I support @Scann & @Avocadobabygirl's opinions here. Winnieswikiworld (talk) 22:31, 24 February 2026 (UTC)reply
    I absolutely support these opinions, and added a bit more of practical experience on this below, and also how I think that this criteria can backfire and lead to more commercial self-promotion. Adaniel (talk) 08:33, 1 March 2026 (UTC)reply
  •  Oppose I don't really see this as an improvement. There should be a clear statement of what each of the notability criteria refer to. To have a set of guidelines that starts with "Note that the guidelines below are intentionally left a bit vague." is not good enough.
My thoughts on Criteria 1: I always understood that Criteria 1 was used for interlanguage links, i.e. to link equivalent articles across different language Wikipedias, if that's correct then shouldn't Wikimedia Commons be removed from Criteria 1 as it's a monolingual project? Also, I've come across some discussion that "small" Wikipedias can be abused by spammers and so Wikidata items shouldn't be automatically created from these.
Criteria 2: I think this needs to be tightened up, something along the lines of "An item passes criteria 2 if it's the subject of an encyclopaedic article, biographical article, newspaper article, journal article published in a widely recognised encyclopaedia, national newspaper, journal. Being "widely recognised" would require at least one Wikipedia article about the encyclopaedia etc.
Criteria 3: From https://www.wikidata.org/wiki/Help:About_data "Wikidata provides structure for all the information stored in Wikipedia, and on the other Wikimedia projects. [...] All data stored on Wikidata can be used to generate all kinds of automated and up to date lists or tables or other structured pages in any Wikimedia site or elsewhere." The way I interpret this is that any named entity contained within a Wikipedia article (or in another encyclopaedic article etc.) would be eligible to pass Criteria 3, for example a person's parents or school.
External IDs: are a mess. Why are social media sites and genealogical sites mixed in with encyclopaedias? Why do VIAF cluster ID (P214) and ISNI (P213) sit at the top of the External Identifiers list when they're both Wikidata property for an identifier that does not imply notability (Q62589320)?.
References: An item shouldn't be allowed to be created without a good reference. imported from Wikimedia project (P143) shouldn't be allowed. The source of a reference should have a stated in (P248), published in (P1433) qualifier (or equivalent) which must pass Criteria 1 or Criteria 2.
Other: It should be made clear how Wikidata is used by external projects and what these projects are and how that relates to notability. Especially concerning GLAMs (Galleries, Libraries, Archives and Museums). Piecesofuk (talk) 10:33, 26 February 2026 (UTC)reply
Criteria 1 is not just about Wikipedia and the other multilingual projects. It is there to address all sitelinks that Wikidata has to other Wikimedia projects, including Commons.
For criteria 2 it seems like based on the rest of the discussion here tightening it even further as you are suggesting is not going to get consensus. Lydia Pintscher (WMDE) (talk) 09:52, 22 March 2026 (UTC)reply
  • I understand why the technical scaling issues motivate a notability policy, but why is a notability policy needed for the social scaling issues? Wouldn't it work just as well to add a new boolean-valued notability property? Then users of the data who want only the higher quality items can restrict to notable items, and other users can search all items. Personally, I'm interested in creative works whether they're notable or not, and I'm fine with lower quality data about the non-notable creative works. dseomn (talk) 20:54, 27 February 2026 (UTC)reply
    Because I believe Wikidata is a project that only works and thrives with its community. The more SEO spam, self-promotion, etc we have the less time and attention we can spend on the core of what Wikidata is meant to provide the world: basic general-purpose data about the world. A new boolean-valued notability property I fear will just shift the debate, because we will disagree on which Item should be classified how. Lydia Pintscher (WMDE) (talk) 09:55, 22 March 2026 (UTC)reply
  •  Oppose

I have been working on GLAM projects in a decolonisation context, bringing collections online and helping surface forgotten, omitted, or unidentified Indigenous people from institutional holdings. In parallel, I have worked on music-industry identifier policies outside Wikidata. From both perspectives, I find the new requirement of having at least one external identifier extremely problematic. The music industry is dealing with similar reforms, and on similar quantities of nodes like Wikidata, and the proposed changed are far more gradual. In music, currently about half a billion sound recordings are available on various platforms, and there are about 2 billion entities that must be connected to them with PIDs to ensure that the music can be legally used and paid. And while there is money in it, the introduction is far more gradual because of the lack of access to registering PIDs.

Over the past 5–6 years, Wikidata has done something remarkable: it has de facto provided a low-barrier, free persistent identifier through QIDs. These QIDs often serve as public nodes where knowledge can start to accumulate, long before a person, object, or community receives a formal PID elsewhere. In many cases, the Wikidata item is precisely what later enables proper identification and the issuing of an external PID. A Wikidata QID is like a condensation point, a particle in the air, that will later attract enough water particles to make a raindrop.

Maintaining PIDs is expensive. The practical entropy of VIAF alone demonstrates that cheap, accessible authority control outside Wikidata is often illusory. If you live in an affluent and inclusive country like Sweden or the Netherlands, identifying, for example, colonial informants or marginalised historical actors may quickly lead to a national GLAM institution issuing a PID. But even in much of Europe this is not the case, and even in the EU I know of countries where access to PIDs via GLAM institutes (far the most frequent case) is politicisied and discriminatory. PIDs are typically issued according to notability criteria defined by state or commercial actors. Entire communities fall outside these systems. On the other hand, talking about self-promotion, PR agencies will quickly learn that all they need for a QID is to buy an ISNI and use it for verification.

There is a growing body of literature showing how Wikidata has democratised access to authority infrastructures.

A think that the sunset criteria and the quality criteria should be much reworked. For example, any item that, after three years, has not resulted in high-quality statements or an alternative PID maintained outside Wikidata could be deleted. But in many decolonisation and community-driven contexts, years are needed between the moment new knowledge surfaces — perhaps through a researcher or a Commons upload — and the identification of corroborating sources, claims, and structured statements. And this is not one day.

I would also welcome a more dedicated, organised space for highlighting new knowledge elements that require corroboration. That would strengthen quality control without discouraging early-stage documentation.

In my experience, many problematic items were created by half-baked bots or mass imports from systems like VIAF — which itself is currently collapsing under inconsistencies and outdated data. Or by people without any experience on editing Wikidata.

Quality indicators could be more nuanced than simply requiring a sitelink or an external PID. For example: the presence of multilingual descriptions, a minimum number of coherent and well-modelled properties, or evidence of structured use across statements. These are better indicators of whether an item is becoming a real node of knowledge.

It would be better to have policies on curatorial policies not on entries themselves. Entries created by human editors who have a track record of having created high-quality entries in the past should be trusted with their judgement. And bots should be much more thoroughly vetted. I think that poor curatorial workflows result in poor quality items, and not self-promotion or the lack of notability of the subject. Who are we to say that a person surfacing from marginalised part of society is not significant? We can however say that a page has no prospect of promoting the accumulation of open knowledge, or that an editor/bot is not yet comopetent enough to create good pages.

New entries should be encouraged — but paired with strict and realistic expectations that they mature into properly structured knowledge nodes. As it stands, the proposed policy feels much stricter than many identifier regimes in the music industry, where billions of PIDs determine whose music is heard and paid out, and where there is a monetary incentive to improve the data.

I will try to join the roundtables, although unfortunately the proposed times are quite challenging for me.

08:15, 1 March 2026 (UTC)Daniel Antal

  •  Oppose as per Scann above.--So9q (talk) 00:40, 3 March 2026 (UTC)reply
  • "Routine listings, directory entries or being quoted in an interview about another topic do not constitute significant coverage on their own" - A number of historical people, places are known little other than their name. So I believe we should limit this term to contemporary entities. (We can consider entities that is only described in any source that is published in last 95 years as contemporary.)--GZWDer (talk) 16:26, 4 March 2026 (UTC)reply
    +1 So9q (talk) 18:30, 4 March 2026 (UTC)reply
  •  Support: I've had the (mis)fortune of seeing first-hand the abuse of Wikidata perpetuated by factions who want to do nothing but promote themselves and their products. Whyiseverythingalreadyused (t · c · he/him) 01:34, 9 March 2026 (UTC)reply
  •  Oppose While I definitely support improving the vagaries of criteria #2 and keeping commercial and self-promotional interests out of Wikidata, notability policy revisions must be careful not to conflate these interests with those of good-faith knowledge workers, organizations, and projects that seek to improve public knowledge for the public good. We must take into account the many WikiProjects that seek to increase the visibility of marginalized and/or under-resourced individuals and communities. Wikidata is not an encyclopedia...applying stringent notability guidelines similar to those of Wikipedia could serve to increase existing bias and privilege. Instructions to simply create Wikibase instances ignore the limits that this potentially imposes on organic, global collaboration, as well as the limited time and labor resources that many must contend with. More nuance in “knowledge equity” could greatly help to mitigate these concerns.Katburl (talk) 18:18, 10 March 2026 (UTC)reply
  •  Oppose I understand that there's a growth related issue with Wikidata, and that there's nothing that can grow forever without problems (not even the Universe). However, the proposal won't solve the issue (there will still be possibility of endless growth) and will create problems for lots of institutions that are starting now to understand that Wikidata is the place to be. I'm working on projects that will have the idea of linking their collections with Wikidata, and many others are doing the same. The same applies, as explained above by others, to scientific institutions, journals, libraries... the proposed change would make that not possible, or possible and then the policy not implemented (or, even worse, randomly implemented). Are there other solutions? Indeed, there are: let's think that I want to add a bunch of humans. I add the instance of (P31):human (Q5) identifier and Wikidata knows that this should be stored in Wikidata/Humans. The same applies to all the WikiCite items: Wikidata/Journals is their place. Wikidata/Taxon, Wikidata/Lexemes, Wikidata/Geo... whatever it's neede to separate entities into buckets when those buckets are large enough... and then federate the query if needed. Of course, there are other solutions, but let's think on how this affects communities, institutions and our reputation, and if the solution is even enforceable or valid in 5-6 years. -Theklan (talk) 08:46, 11 March 2026 (UTC)reply
    The idea sounds intriguing but people are already asking for the scientific article split to be undone. And that is a split that is relatively straight-forward. Splitting out humans for example would have much more far-reaching consequences.
    And it is ultimately not just about the query service, where this could be split. It is also about the amount of edits in Wikidata for example, which would not change here. Lydia Pintscher (WMDE) (talk) 09:58, 22 March 2026 (UTC)reply
  •  Oppose I agree with Scann and Katburl's comments above. I strongly prefer technical solutions to technical problems, and the proposed changes to Wikidata's notability policy will do more harm than good. I would be in support of better multilingual community outreach for Requests for Comment in general--it doesn't seem fair to have decisions like this happen only in English in a corner of Wikidata that most everyday editors never notice or see. Decisions made in echo chambers do not serve the entire community. Wikidata is frequently the first identifier to be assigned to notable people, institutions, and ideas. If we need to wait for closed data sources to make an identifier for something first, the usefulness of Wikidata is diminished (in my mind, and in my current Wikidata practices in the GLAM environment). --Crystal Yragui, University of Washington Libraries (talk) 22:12, 11 March 2026 (UTC)reply
    Technical solutions are of course part of what needs to happen and is happening. But we also need policy solutions. They have to go hand in hand.
    About more outreach: People have done outreach for example in the Latin-American communities to gather input there. And I welcome more. I'm happy to support anyone who does. But ultimately an RfC is how the Wikidata community needs to make this decision. Lydia Pintscher (WMDE) (talk) 10:00, 22 March 2026 (UTC)reply
  •  Oppose I can't support this in its current form. In the last year, I've gotten to know the archivist community well enough to learn what a priceless gift the current policy is to free knowledge workers. Let us preserve artisanal Wikidata production, especially for knowledge equity topics, even as some extra additional care might need to be taken around mass-editing policy. In addition, I don't think it has ever been clearly articulated what non-"serious" sources we are looking to restrict, and if the actual uses would make a technical difference; for example, do we mean video game databases?--Pharos (talk) 19:24, 16 March 2026 (UTC)reply
    I understand the sentiment. But the current status isn't sustainable unfortunately. We do need a new policy. Do you have suggestions for how to change the current draft to make it better?
    About non-serious sources: For example yet-another-crypto-company creating Items for them and their products based on an article in a crypto-advertising magazine. Unfortunately there is quite a bit more on WD:RFD. Lydia Pintscher (WMDE) (talk) 10:02, 22 March 2026 (UTC)reply
    Surely we can treat nonprofit/academic efforts differently from commercial/promotional ones? Would you be amenable to a reformulation of the policy that focused on that distinction? Pharos (talk) 16:21, 23 March 2026 (UTC)reply
    I am clearly against this two-class system  Oppose --Gymnicus (talk) 16:34, 23 March 2026 (UTC)reply
  •  Comment about the actionability of the policy (also the present one, BTW): I also report here phab:T420316 (adding the possibility to delete items through the REST API); if solved, it will also make possible to add to QS 3.0 the possibility of mass-deleting lists of items (cf. meta:Talk:QuickStatements 3.0#Feature: delete items). --Epìdosis 10:21, 22 March 2026 (UTC)reply
    I have also just requested an admin-bot periodically deleting surely non-notable items, as defined by the current WD:N. Epìdosis 17:07, 21 April 2026 (UTC)reply
  •  Oppose The current process is sustainable until and unless we establish a more meanful, participatory, community consensus to change it. Also, we are switching to a new SPARQL engine in June 2026 to be finalized summer 2027, in the context that we have been using our current technology with no updates since 2015. Our culture will change so much with expanded technology that the familiar limits which we have needed in recent years do not make sense in the context of greatly expanded, near-future capacity updates.
This RfC is a great start to orienting that discussion, but this RfC is not a representative discussion to make what would likely be a 10-year new strategy for the fundamental nature of Wikidata. The proposed draft changes would significantly disrupt several of the largest and most popular Wikidata community editing efforts, including Wikidata:WikiCite; efforts to catalog cultural collections including Wikidata:WikiProject sum of all paintings and Wiki Loves Monuments (Q1353202); efforts to create biographical entries which Wikipedia community projects have done for decades by occupation, demographic, and activist need; and the next-generational institutional collaborations with universities, government agencies, and cultural and research organizations which always have been around but which are growing especially now in the context of greater interest in data science. As an alternative to this RfC to refine our rules, the counterproposal that I would make is to determine what datasets we want to import then design rules to match that, rather than design rules before checking to see what they can import. Before I support rule changes, I would like to get straightforward confirmation that existing Wikidata community projects will remain welcome and encouraged. The text in this RfC seems like it could be top-down disruption of a lot of community organizing, and also I am not seeing much organized community discussion in this RfC.
I see some other people have already stated these points, but I will state them in my own words -
  • We should not require public database identifiers as an inclusion criteria. I get the benefit that importing identifiers is super-popular in Wikidata and agree that we should do it, but establishing identifiers is the game of the privileged. Having an identifier does not make something notable, it just means that a thing is more connected to the bias of institutions which mint identifiers. I do not want to either 1) further solidify through Wikidata that concepts matched to external identifiers are more real and worthy or 2) further diminish the value of concepts without identifiers. Worthy sets of items without identifiers include collections at underfunded universities, museums, government agencies, and nonprofits, which includes most of those kinds of organizations in the world. I can compromise: we could have one set of rules which applies to institutions which are high-economic development or commercial; and another set of rules which applies to institutions which are low-economic development or university/nonprofit/research/cultural.
  • We do not have a Wikidata definition for reliable sources, so we should not enforce a requirement for sources until we establish such a policy. I agree that Wikidata content should cite reliable sources, but before we require that kind of fact-checking, we need a policy definition on reliable sources, and we do not have one. We have Help:Sources which is not a definition and not what we need. I oppose establishing a requirement for fact-checking, in the context that we have so much disagreement on how to do fact-checking or what constitutes a reliable source. We cannot import English Wikipedia's reliable source policy as some are suggesting, for reasons including that Wikidata needs to cite some fundamental databases which are reliable in senses and different ways than English Wikipedia can evaluate. Now is not the place to discuss this, but convening a task force to draft a Wikidata reliable sources policy would be useful.
  • Prohibit commercial marketing and branding; allow similar activities from universities, cultural/research orgs, and government agencies I oppose Wikidata:Self-promotion. At English Wikipedia, there have always been debates about COI/conflict of interest editing. English Wikipedia has tens of thousands of COI complaints, and I know there is a lot of confusion about this, but those complaints are almost all commercial marketing and branding. We have had trouble writing out COI rules, but the reality on the ground is that some COI editing is good, and some is bad. The good kind is for example museums importing data for their collections into Wikidata, or other comparable efforts to get popular and user-requested info into Wikidata. The bad kind, which is 10,000x more common than the good kind, is promoting brands and products. The custom for writing Wikipedia/Wikidata rules is to present extremely short rubrics intended to apply to all knowledge in the world. While short rules make it easy to on-board new editors, they also fail by, for example, differentiating crypto scams designed to cheat Wikipedia from valid and wanted entries for the knowledge collections of cultural institutions. I would like our COI/self-promotion policy to start with intent to support Wikidata editing communities, and to avoid leading with brief rubrics that could constrain the human editorial process here. Let's recognize and welcome institutional partners who have data collections of value, and let's exclude commercial and marketing interests which scammers continually redesign to try to sneak through our rules. I think a better start to our COI policy would be to keep it for humans, by humans, who can readily identify the difference between commercial interests and not for profit knowledge sharing even when a short rule cannot.
  • Let's remerge the scholarly graph into the Wikidata main graph! WikiCite is the most popular Wikidata project in terms of institutional partners, external funding invested into the Wikidata platform, count of university staff and librarians contributing, and the single-largest Wikidata community editorial user project. There was a recent Wikidata graph split which made sense in the context of Blazegraph, the SPARQL backend, being at capacity. Wikidata is about to go through a database migration, and after that migration, the benchmarking data indicate that Wikidata has capacity to remerge and complete the (finite, completable) Wikicite project. Wikicite is our best-chance best-idea effort to have a strong connection between the Wikimedia platform and every university and researcher in the world. We need this knowledge and fact-checking, and designing Wikidata's inclusion criteria and other policies to advance Wikicite is a proven high-impact strategic direction to achieve Wikimedia Movement goals. To me, the worst part about the Wikicite graph split was stumbling around for years in Wikidata, finding that we lack community discussion and leadership infrastructure to even talk about major problems when we have them. I wrote about this in "Wikidata Graph Split and how we address major challenges", February 2026 in The Signpost.
I am not persuaded that we should make decisions in crisis right now. We have years to talk this through and there is no urgency. We just did a graph split of Wikidata. Also there is a migration from the Blazegraph backend to some other database backend, which should expand the capacity of Wikidata's SPARQL engine at least several times, which further removes urgency.
Bluerasberry (talk) 17:41, 6 April 2026 (UTC)reply
We do not have years to make this decision and the graph split is not going to be undone. The new Query Service backend will help but it does not address any of the other reasons why we need to look at the Notability policy such as admin overload and other technical limitations such as the amount of edits. Lydia Pintscher (WMDE) (talk) 15:24, 8 May 2026 (UTC)reply
  •  Oppose That something has to exist first in another part of the wikifamily is wrong and says that wikidata is not an equal part the family, if I was for example going to write an artical about a local politician in wikipedia, I would start by gathering the facts first in structured way in wikidata before writing it out long-hand as an article. It is also a dangerous path to put limits on how people work/think, I am not unusual as an IT engineer to think about the data/database before working on the "frontend" and it as a valid a way as someone who would start with "freetext" first Back ache (talk) 13:43, 13 April 2026 (UTC)reply
    @Back ache: Where in the notability criteria is this stated? Nowhere, because being included in another Wikimedia project is just one of three way to achieve notability. So please don't spread such misinformation. Gymnicus (talk) 14:23, 13 April 2026 (UTC)reply
  •  Comment I realise Wikidata is facing some difficult choices at this point. I just want to weigh in with another GLAM partnership perspective. In my advocacy of Wikidata, one of the things I'm most excited about is Sum of All Paintings: the idea that Wikidata might point us to all the works by a notable artist, all the art works from different cultures that depict a particular animal or are associated with a particular location. Of course there is more to visual art than paintings and using Wikidata to aggregate other collections is just as inspiring. Such data sets can help with Wikipedia lists such as w:en:Fabergé_egg#List_of_eggs or w:en:Dala'il_al-Khayrat#Manuscripts, even when the individual objects do not merit a Wikipedia article. "Self-portraits of women" is a query that I use in all my Wikidata training; I talk about how assembling such an image gallery from hundreds of museum or gallery websites would take a ridiculous amount of work.
In my work as a Wikimedian In Residence, I have been involved in multiple activities that the proposed changes might rule out (if I understand the proposals correctly):
Cultural institutions, even the best-known ones, are often poorly served for structured data: they might have a web site and print publications but not always a database. Most of their collections' art works lack third-party coverage or their own Wikipedia article, but they are documented by experts and usually published in a form that is regarded as authoritative, in collection catalogues or exhibition catalogues. Most relevantly, lists of items are incredibly useful open knowledge about artists' careers, evolution of styles over time, or different perspectives on a theme.
Is Wikidata saying that it's no longer our ambition to have every painting/ manuscript/ astrolabe/ painted-porcelain-bowl-by-notable-artist in a GLAM collection? Are they to be left to their own Wikimedia project (we can't rely on the cultural institutions themselves to create the aggregators, although they do appreciate them)? Can I create an item for an object to be used in the Commons page for the object's photograph, without having a Commons sitelink? MartinPoulter (talk) 14:58, 13 April 2026 (UTC)reply
As a relative newbie to Wikidata editing, I completely agree with Martin's comments. One of the troubles notability criteria creates for archive material in particular is that often the catalogue metadata about the archive is the only source / evidence that something exists, and peer-reviewed, verifiable works about collections take huge amounts of time to create - it's not something within the scope of most of us working in the sector. However Wikidata has the potential to connect these collections to wider knowledge, which just wouldn't happen anywhere else: it's a much more equitable way of joining up knowledge. I worked on a small project last year which explored the data held in my institution's College Periodicals to Wikidata: as a collection, this material is unexplored and little has been published on it. As Wikidata, there is information in there which can connect 20th century London life to all kinds of knowledge. Under the new criteria this project wouldn't go ahead, and neither would a piece of work I'm looking at undertaking this summer creating Wikidata entries for the UNESCO-registered George Orwell Archive. We constantly get news articles saying how 'undiscovered' these archives are when they are actually fully digitised and accessible; using Wikidata as a test case helps us to see what's feasible with these collections, what isn't and where the best place to concentrate Wikimedian in Residence efforts is. Octavosaurus (talk) 16:00, 13 April 2026 (UTC)reply
I also agree with @MartinPoulter's comments. One of the huge benefits I see to Wikidata is the ability to make unexpected connections between entities and create that linked network of data, discovering connections that would otherwise remain hidden. I have worked on 2 small projects so far - one with theses metadata and one with suffrage interviews metadata. Both make important contributions to the scholarly graph, and reveal connections and relationships which would otherwise remain hidden. Under the new criteria we would be hindered in making this connect accessible and both content and connections would be less accessible to researchers - or entirely inaccessible. HelsKRW (talk) 09:01, 14 April 2026 (UTC)reply
Thank you, @MartinPoulter. The examples are helpful. Do I understand you correctly that one problem is that for some of them you can point to sources but they are not external IDs in the strict Wikidata sense? This I think we can try to find solutions for. But the second one it sounds like Wikidata is the first place they are really described. And that seems quite at odds with Wikidata as a secondary database. And that seems quite important to me as a Wikimedia project. We're not doing original research. That is part of why I believe the Wikibase Ecosystem, and Wikibase Cloud in particular, is so important so that the primary sources can be built up. Lydia Pintscher (WMDE) (talk) 15:32, 8 May 2026 (UTC)reply
Sorry @Lydia Pintscher (WMDE) that I didn't see this reply at the time. "for some of them you can point to sources but they are not external IDs in the strict Wikidata sense" - yes, this is exactly what I meant. I agree that there has to be an authoritative source of the information outside Wikidata and that Wikidata is not a platform for original research. All the statements I mention are previously published, just not on an RDF platform. They are on the web sites of the museums or sometimes in printed catalogues. With the aggregated astrolabe dataset, Wikidata allows us to ask questions like "When was the biggest astrolabe made?", "Where are the astrolabes with Hebrew text?" "How many astrolabes in public collections were made in the 15th century?" Not all of these individual objects is going to have a Wikimedia sitelink but they have all been described by curators, usually given their own web page, and have a catalogue number. MartinPoulter (talk) 16:25, 1 June 2026 (UTC)reply

 Comment I agree with concerns above about the risk of excluding useful and reliable GLAM information based on the currently proposed wording of the criteria, and with @Daniel Antal's statement that there are alternative notability indicators, such as coherent properties and well-structured statements and references, which are more likely to make something a useful node than having a single external identifier from limited "recognised" places. In terms of references, which I broadly support us in encouraging, the wording is currently unclear. It is not clear what constitutes "significant", "serious" or "publicly available": I think questions such as the following are not clear to me from the current wording: What is required to indicate "significant"? Some people (e.g. matrons who head all nursing services in an organisation) were not written about at length in any single source, but their importance and interconnection with other data (e.g. historic hospitals) are evident from short but numerous references to their work. In terms of "seriousness", would a publication with a popular audience, such as a general-audience gardening magazine, be sufficient as a reference for the biography of a specialist plantperson responsible for discoveries of species and breeding of hybrids? Is the extensive database Lost Hospitals of London, widely respected by medical historians, to be taken seriously? And then the availability: if a source is not digitised but exists as a physical item, in a single physical repository where it is available to visitors coming to that place to consult, does it count as "available"? What about if it is digital but paywalled or only accessible to those in a specific geographical IP range - is that "publicly" available? Zeromonk (talk) 11:29, 14 April 2026 (UTC)reply

To bring another GLAM perspective: I worked for the UK's largest local museum service, which had no publicly available catalogue (due to the design of their IT system). I feel this is pertinent to the discussion, as parts of the collection were designated internationally significant (but not neccessarily published). It seems under the current proposal, these objects would be excluded from Wikidata, unless in a secondary catalogue somewhere? This seems to run against the "sum of all human knowledge" goal Lajmmoore (talk) 18:27, 20 April 2026 (UTC)reply
@Lajmmoore Thanks. For me an important point is that "sum of all human knowledge" is and will remain relevant and a goal to work towards. The questions for me is just where. And I'd like us to move to a world that truly embraces the idea behind linked open data in a more distributed network instead of putting all our eggs in Wikidata. Lydia Pintscher (WMDE) (talk) 15:38, 8 May 2026 (UTC)reply
  •  Comment I am coming to this discussion as a Wikidata contributor, but primarily as someone working at a Wikimedia affiliate. Many of our projects are collaborations with cultural heritage and higher education institutions, and Wikidata is a significant part of many of those projects. For previous work see Scottish witch trials data, the Atlas of Hillforts, or efforts to share metadata about doctoral theses. All of these have enriched Wikidata directly; benefitted the data shared through processes of data cleansing, reconciliation, and enhancement; and have encouraged other projects to engage with Wikidata.
I think that the proposed changes would create a chilling effect that would discourage GLAMs, researchers, and higher education institutions from engaging with Wikidata. Specifically, it may be unclear what counts as a self-assigned ID, and the guidance around conflicts of interest is very broad.
I appreciate that the scale of Wikidata poses technical challenges – and indeed maintenance challenges as I have noticed that vandalism can persist for longer than ideal – and therefore there needs to be consideration of what constraints are put in place when it comes to what can be covered by Wikidata. At the same time, I came here to voice concern about the possible impact of the revisions to the policy. The proposal would create two tiers of external identifiers but the distinction will not be immediately clear. When identifiers are designated by a research project would that count as ‘an independent authority’ or are they ‘self-assigned’ by the researchers? As a general rule, institutions do not want to run afoul of Wikimedia policies and so will tread carefully. If the policy introduces uncertainty or unclarity that will create situations where data holders are less likely to engage.
The conflict of interest issue should really be treated separately from notability – and it is counterintuitive for a policy to hold up an essay as that has not yet been elevated to policy level – but beyond that is also potentially an issue for affiliate work. If an organisation wishes to share data is it prevented from doing so? For example, if that had been in place as written would it have prevented the witch trial data being shared by the University of Edinburgh? There are two considerations: what we think should happen as Wikimedians, and what a more external audience would have understood the policy to mean. What may appear clear to us as Wikimedians may raise questions amongst potential partner organisations. I am concerned that this would have an inadvertent chilling effect and would be significantly detrimental to Wikidata in particular, and the Wikimedia movement as a whole. Richard Nevell (WMUK) (talk) 09:36, 30 April 2026 (UTC)reply
@Richard Nevell (WMUK): Thank you. The point you're raising about the intent of the policy vs. how it will be read is an important one. I'll send you a message so we can maybe dig into this a bit more in depth. Lydia Pintscher (WMDE) (talk) 15:41, 8 May 2026 (UTC)reply
  • I think stronger criterias should be implemented specifically for living persons and current organisations (especially companies); this would mitigate most of the autopromo problem. Jules* (talk) 17:03, 27 May 2026 (UTC)reply
    I am opposed to this. It makes no sense to establish a two-tier society—on one side, the "bad" living persons and current organizations, and on the other, the "good" GLAMs. To me, that is the wrong path. --Gymnicus (talk) 10:22, 30 May 2026 (UTC)reply
  • I like the new wording for identifiers and sources. The proposed phrasing seems to bring them in line with common practice at RFD and AN (requests for undeletion). I have concerns about the change from "AND" to "OR". This seems by far the most significant part of the proposal, and we need to be clear about what we're getting into here. For me, N2 comes down to maintainability. For Wikidata to be successful, it needs to be possible for editors in general to be able to maintain items without recourse to the original item creator. Maintainability relies on both identity (knowing when to merge or hold separate) and sources (allowing us to verify, update, and extend). Ensuring both of these is why N2 has an "AND". Regarding GLAM and archival records, do we have any examples of items that lack all of sitelinks, identifiers and sources, and structural need, but which should nevertheless be notable because of an archival record? Bovlb (talk) 17:13, 8 June 2026 (UTC)reply


Eligible wikis

[edit]

Wiktionary

[edit]

The proposal like the old policy Wikidata:Notability both lack wiktionary in the long list Wikipedia, Wikivoyage, Wikisource, Wikiquote, Wikinews, Wikibooks, Wikidata, Wikispecies, Wikiversity, or Wikimedia Commons. Wiktionary should be listed although certain namespaces including ns ZERO are excluded. Templates, help pages, appendix pages, etc ultimately should be fully accepted. Taylor 49 (talk) 19:04, 4 March 2026 (UTC)reply

 Support Empat Tilda (talk) 18:17, 7 March 2026 (UTC)reply
 Support: sounds good to me Whyiseverythingalreadyused (t · c · he/him) 01:34, 9 March 2026 (UTC)reply
 Support and also support for all other WMF content wikis, current and future; the "long list" is unnecessary.--Pharos (talk) 01:56, 31 March 2026 (UTC)reply
 Support dringsim 14:08, 6 April 2026 (UTC)reply

Non-WMF wikis

[edit]

There should be a subpoint clarifying that technical details of wikis (category, template, modul, namespace, ...) are in scope ONLY for WMF wikis. Taylor 49 (talk) 19:04, 4 March 2026 (UTC)reply

Since this is under criteria 1, which is about the Wikimedia projects I believe this is already sufficiently addressed. Lydia Pintscher (WMDE) (talk) 11:47, 5 March 2026 (UTC)reply

GLAM topics on Outreach-Wiki

[edit]

Somewhat parallel to the marginalized knowledge discussion above, I have made a proposal for a representation of GLAM topics on Outreach-Wiki.--Pharos (talk) 17:37, 13 April 2026 (UTC)reply

Round 2: Make Wikidata:Self-promotion a policy

[edit]
Discussion about making the essay about self-promotion a policy. It seems at this time the answer is no. For the new policy proposal the wording was tweaked to take the input into account.

Wikidata:Self-promotion is currently an essay. To be referenced in the new policy it would need to be elevated to policy. Do you agree with making it a policy?

Votes
  •  Oppose as per Fuzheado below and Scann above (banner on main page or better organized public vote like the one in Abstract Wikipedia perhaps?). We have 16k editors and only a handful has participated so far. Compare that to the AW vote.--So9q (talk) 00:45, 3 March 2026 (UTC)reply
What vote has there been around policies for Abstract Wikipedia? Ainali (talk) 06:46, 3 March 2026 (UTC)reply
there was a naming vote recently. They advertised it and had a lot of participation compared to the number of active editors I would say. So9q (talk) 18:24, 4 March 2026 (UTC)reply
  •  Support: see WD:RFD and the self-promotion over there Whyiseverythingalreadyused (t · c · he/him) 01:35, 9 March 2026 (UTC)reply
  •  Oppose It is hard to make a brief rule which applies to all cases. Almost all of the problems come from commercial entities or individuals doing marketing and promotion. This draft policy is too close to preventing universities, cultural organizations, research institutions, and government agencies from sharing general-reference datasets about their knowledge collections. I think a self-promotion or COI policy should be written to practically prevent marketing, which is 99% of the problem, but also invite certain classes of knowledge organizations to share their collections. Bluerasberry (talk) 15:58, 7 April 2026 (UTC)reply
  •  Oppose very bad rules on self-promotion (and their translations/interpretations) have already been weaponized on smaller/medium Wikipedia instances to suppress minoritized/progressive content from civil society organizations by conservative contributors/admins (this remains an issue though it has been recognized for over a decade). Institutions and Corporations have more capacity to strategize visibility (and marketing) without doing simple self-promotion and in-balance in coverage will stay if not just make it even harder for less-resourced. Zblace (talk) 08:18, 19 April 2026 (UTC)reply
Discussion
  • There is a slight mismatch between the essay and the proposal when it comes to the importance of external identifier. It says "A listing in a business registry or government database may act as an identifier, but it does not by itself make a company notable. Independent coverage in reliable sources is required to firmly establish notability." which is not exactly what the proposal is trying to say with 2.1. The solution is either to soften the tone in the essay, or revert the addition of the "OR" for point 2.1 and 2.2. If they both are required, it is more aligned with the essay. Ainali (talk) 18:42, 19 February 2026 (UTC)reply
  • I'd say yes to approve the self-promotion policy, but not without a previous discussion focused on it. Maybe it is for another time, but I agree that it should happen sooner than later. Another option to make the mention coherent in this draft for the Notability policy, is to change "not permitted" for "is strongly discouraged", which is actually in line with Wikidata:Self-promotion. Mariana Fossatti (WK?) (talk) 14:16, 20 February 2026 (UTC)reply
    I concur with the idea to soften the language – currently, "not permitted" is too strong. While I fully support the goal of protecting Wikidata’s neutrality, I believe the current proposal of a hard ban is strategically problematic and out of step with our community norms. I suggest we instead use "strongly discouraged" and move toward a principle-based definition for the following reasons:
    • Consistency with sister projects: Aligning with established norms like en:WP:COI (also using "strongly discouraged") ensures that contributors moving between projects don’t face contradictory rules. A "not permitted" stance on Wikidata would create confusion for cross-wiki contributors.
    • Value human judgment: Hard bans are often blunt instruments. They risk punishing good-faith contributors who may be the most qualified to provide technical or verifiable data (such as working with GLAM institutions or professional organizations). We should prioritize community oversight and human judgment over rigid prohibitions.
    • Returning to a "prime directive": Rather than relying on an arbitrary list with "close relatives" or "other entities," we should adopt the principle-based language that once served Wikipedia so well on its COI page:
    "When advancing outside interests is more important to an editor than advancing the aims of Wikidata, that editor stands in a conflict of interest."
    This framing provides a clear, ethical North Star. It focuses on the behavior and intent of the contributor rather than just their identity, which is far more useful for long-term community health.
    - Fuzheado (talk) 19:05, 23 February 2026 (UTC)reply
The problem with making that “OR” into an “AND” is that we don’t (yet) have a lot of external identifiers for many fields, including government-related directories for basic inventions or historical laws. This is not because those directories don’t exist, but mostly because they are still not digitized further back than the 1990s or so. Jane023 (talk) 11:18, 21 February 2026 (UTC)reply
  • Wikidata should have a clear COI policy, but the current essay has some real problems. Among them, it assumes as a given that only non-notable entries will be added, that the notability of an item is determinable by who added it, that the COI contributor doesn't have sources, and it is written like a user essay, not a site policy. Compare to the existing policy on paid editing, which is much stronger.
  • For the Notability policy draft itself, I would recommend omitting the following: "Creating items about yourself, your close relatives, your business and other entities you have a conflict of interest with is not permitted." The identity of an item's creator should not determine the item's notability—Notability is about scope, COI is about behavior. Otherwise, my second choice would be changing "not permitted" to "strongly discouraged". WWB Too (talk) 20:19, 23 February 2026 (UTC)reply
  • This would need to be worded differently, otherwise it would have serious negative consequences for Wikimedians in Residence (like me), who are in a way a special case in COI but their and their institutions' contributions are important to Wikidata. Also, I don't think a general ban is a good idea, rather something similar to Wikipedia's COI policy, which allows COI editing if that is fully transparent and complies with all other policies.
On a more theoretical level, I'm not sure if I understand the idea of COI when it comes to data. On Wikipedia, having a COI assumes that you are more likely to introduce bias in what you write about and especially in how you write about your subject. In Wikidata, it would be more difficult to overemphasise/overrepresent/distort something, because there is no tone, and adding new items or new statements won't take space from others. I can see how COI could affect what you create new data on but a good notability policy covers that, and if the item is notable, it doesn't matter too much who creates it. --Adam Harangozó (talk) 17:12, 24 February 2026 (UTC)reply
I absolutely agree with the questioning of COI. Creating a page, a QID is about identification. Who is the best qualified to identify a person than the person itself? Moreover, in Europe where we have GDPR, it could be argued that only the persons themselves, or their authorized representatives can legally make an entry. In Europe, if we introduce COI, the we also have to make and opt-in, opt-out form for every living person. Generally, an informed discussion (for Europe) about GDPR and Wikidata would be very useful, because it would not be absurd to make the claim that in Europe making an entry of a living person without explicit consent from the person is illegal. Adaniel (talk) 08:45, 1 March 2026 (UTC)reply
I'd like to back up and support whast @Adam Harangozó has said here: as a fellow Resident, these are crucial considerations when dealing with collections, expertise and institutional knowledge. EriedgenArc (talk) 09:29, 10 March 2026 (UTC)reply
I agree with the comments made by @Adam Harangozó. I too am a Wikimedian in Residence and am of the opinion that a general ban would not be ideal and may negatively impact contributions I and other WiR folk are undertaking that are aimed at improving Wikidata. I would, however, support a COI policy similar to that of Wikipedia. I also support the above comments made on this topic by @Fuzheado. Ambrosia10 (talk) 19:25, 10 March 2026 (UTC)reply
I agree with the comments of @Adam Harangozó, particularly the point that "if the item is notable, it doesn't matter too much who creates it". I would NOT support a COI policy similar to Wikipedia, as that has the opposite approach. DrThneed (talk) 20:13, 23 March 2026 (UTC)reply
  •  Oppose I think a few projects similar to the ontology task forces with the aim of pruning clearly half-made and useless pages (at least marking them as sub-standard and encouraging improvement) would be necessary to pinpoint the problem at scale. Creating a new bottleneck for new knowledge by human actors seems to me arbitrary when half-baked bots are spaming Wikidata with perhaps far more useless, misleading, or resource eating pages, for example, by mass importing data from poorly maintained scientific repositories or VIAF. Wikidata is far too big to create an objective impression by humans, I would suggest evidence-based discussion on this supported by statistics made from well-designed queries. I am not sure that we understand the extent of the problem of self-promotion versus generally mass-importing low quality data --Daniel Antal (talk) 08:41, 1 March 2026 (UTC)reply
  •  Oppose Someone created an item of myself at Wikidata, and then an article at Wikipedia. I understand that there are COI if I change things there, but factual corrections, or adding an external identifier (and ORCID in my case) is not problematic. I also changed my birthdate which wasn't accurate. I don't think this is a COI, but good faith edits and, even more important, they improve the accuracy of the database. Should we ban anyone to improve their item or one they are related to (their institution, for example)? That would require a large knowledge of who is who, because I can even change those without logging-in, with a temporary account, and if the data added is correct (an ORCID number), no one will revert it. -Theklan (talk) 08:53, 11 March 2026 (UTC)reply
  •  Oppose Wikidata is not Wikipedia, and having a Wikidata item isn't equivalent to self-promotion (I'm most concerned with creating items related to an organization one works for or represents). Said another way, creating an item record for an organization or for oneself is not self-promotion. The use of that item is entirely reliant upon how useful it is, and how many connections one can make to other statements and items in Wikidata. The criteria seems rather low anyway (anyone can make an ORCiD ID for themselves...) Further, in the essay there is only detailed guidelines for not creating an item for oneself but very little about "your organisation or your work" which is where most of my concern falls. I have not felt that people creating items about themselves was a pressing issue... is it? Is there data to support that this is a huge problem? Surely the scaling issues are not caused by individuals making items for themselves alone... I feel like there are more important issues we could be tackling. For the record, I have not made an item for myself and would never do so! Thank you for the RFC. Kind data (talk) 15:52, 11 March 2026 (UTC)reply
    For what it's worth, I think that the suggestion of making WP:Self promotion a policy is rooted in the regular posts on RFD and Project Chat from individuals or businesses who want to improve their search engine optimization by creating a Wikidata item. There are two of them in the Project Chat March 2026 archives already. I think a lot of the concerns that GLAM affiliates are sharing in opposition to making WP:Self promotion a policy might be addressed if there was a link to WP:Curator, a section of the English Wikipedia page on COI, or a section in the essay itself noting an exception for GLAM institutions and their collections. Mcampany (talk) 18:28, 11 March 2026 (UTC)reply

Thank you everyone for the input so far on this. Based on the comments I see the following ways forward:

  1. Soften the tone of the essay and take into account especially Wikimedian in Residence-like set-ups, potentially similar to en:Wikipedia:Conflict of interest#Cultural_sector.
  2. Change the Notability policy draft from "not permitted" to "strongly discouraged".
  3. Change the framing in the Notability policy draft or essay to "When advancing outside interests is more important to an editor than advancing the aims of Wikidata, that editor stands in a conflict of interest."

What is people's preference? Lydia Pintscher (WMDE) (talk) 10:10, 22 March 2026 (UTC)reply

@Bovlb Especially, what is your preference as the author of the essay? Lydia Pintscher (WMDE) (talk) 10:11, 22 March 2026 (UTC)reply
Sorry not to be participating actively, but real-life issues are interfering with my ability to spend time on volunteer projects.
My essay is intended to reflect general sentiments commonly held by the community, not just mine, so I am happy for it to become better aligned. In particular, my essays were written based on the old criteria, not the proposed changes.
I am uneasy about "strongly discouraged" because it is functionally similar to "permitted'. Can admins sanction someone for doing something "strongly discouraged"? If not, then it has no teeth.
The change from "AND" to "OR" in the second criterion is significant, and deserves more discussion as it is a significant weakening from the existing criteria. When patrolling item creation, I will typically leave an item if it has either, but I feel that the goal should remain both.
I am very sensitive to the issue of traditionally under-documented regions (knowledge equity), but I stand by the principle that Wikidata should not be the first place that something is documented. If we can't come up with good identifiers and sources, then we're not creating a maintainable item. Unmaintainable items do not further the project. Bovlb (talk) 20:21, 23 March 2026 (UTC)reply
I expressly support point 2. If a data object for a person, a company, or anything else is notable, then every user should have the opportunity to create this data object, even if the user has a conflict of interest. --Gymnicus (talk) 14:08, 22 March 2026 (UTC)reply
Still very much  Oppose making that a policy, which still looks like a quite out of place Wikipedia policy. I don't see any problem with someone creating their own item or items of people involved in some given activity or project based on ORCID, VIAF or other authorities we regularly use, and it often makes a lot of sense when you are doing some kind of institutional project here. Also, I really can't see how that would be "promotional". In my experience, in some occasions I had to deal with academics wanting to delete their Wikidata item as they thought it would be somehow damaging to them. I never had to deal with the opposite case.-- Darwin Ahoy! 17:46, 19 April 2026 (UTC)reply
@DarwIn: Unfortunately there are quite a lot of cases on the Request for Deletion page, and those are only a small part of what gets caught. Lydia Pintscher (WMDE) (talk) 15:53, 8 May 2026 (UTC)reply
 Oppose I think the same as DarwIn, Wikidata is not Wikipedia. And just because there are example of deletions request (that violation of this non existing rule), we should not ban useful contribution. If this was approved we will need to delete notable entries. Jcubic (talk) 22:25, 15 May 2026 (UTC)reply

Round 2: Marginalized knowledge

[edit]
Discussions related to the impact of a new policy on marginalized knowledge. The biggest outcome is expanding criteria 2 to cover archival records.

There are still remaining concerns around the implications of the draft for marginalized knowledge. This section is to explore ways to improve the draft in this regard.

Suggestions for improving the wording in the draft to reduce the impact on marginalized knowledge while maintaining more clarity about what is notable

[edit]
  • Before getting to generalized wording suggestions, let me suggest a very concrete example from my experience:
Centro de Estudios Puertorriqueños (Q17515347) is a research center of the Puerto Rican diaspora that is based at my university. There are 1,691 people and organizations represented in its archive, and though they don't have unique identifiers, they represent important subjects of scholarly research: https://centroarchives.hunter.cuny.edu/agents. Are these items still a good fit for the future of Wikidata? (I believe that they should be...)--Pharos (talk) 02:03, 25 March 2026 (UTC)reply
It all comes back to maintainability. How can another editor verify that the claims are correct? How could we determine whether two items are duplicates of each other? If we cannot answer these questions, then I submit that the items do not belong in Wikidata. Bovlb (talk) 14:44, 25 March 2026 (UTC)reply
These items are essential to Wikidata’s (and, I would argue, Wikimedia's) future. The items are on a reputable institution's website, and can be linked to, which is already exceptional compared to most of the world's knowledge, and should be amply sufficient. To Bovlb’s point on maintainability: we will only get more maintainers if we become more welcoming. By hosting this data on Wikidata, we show community experts and researchers that they are welcome and belong here.
A general note ... pushing marginalized communities toward Wikibases is ethically problematic; it strongly and actively reinforces existing biases. And just on the practical side: most underrepresented knowledge holders around the world don't have the finances, IT knowledge and critical mass to realistically keep their own Wikibases maintained. Seeing the political reality in most places in the world, that will only become more universal, not less.
By making our platforms less welcoming to diverse data, we alienate the editors needed to maintain it. If we want a global reputation for accuracy, balance and equity, we must invest in growth of our core platforms, and inclusion of data and people, not defensive and selective exclusion. Spinster 💬 19:40, 25 March 2026 (UTC)reply
You do not need to motivate why we would want to include marginalized knowledge here, as the community is generally open for it. What is missing is an actionable, concise definition how to handle such knowledge within Wikidata in practice, as we have little experience and policies for it.
Wikidata:Notability is currently a problematic policy for marginalized knowledge, as it ties notability and therefore inclusion to the presence of external (to Wikidata) sources. This is mainly motivated by verifiability considerations. Without verifiability, we find ourselves in a situation where we cannot answer basic questions regarding a given item such as:
  • What is this thing?
  • Does it exist, after all?
  • Does it belong into Wikidata? Why was it included in Wikidata?
  • Is it the same as some other thing with a similar label?
  • Is the information about this thing on its item page correct?
  • What is a proper definition in Wikidata (i.e. basic set of statements)?
  • … (and so on)
The community at large is only able to work with data items without sources in the classic sense if we find answers to this problem. These answers would also be a sound foundation for the inclusion of marginalized knowledge in Wikidata that cannot be linked against any external sources.
Please also be aware that besides highly welcomed valuable marginalized knowledge, we unfortunately do experience inclusion of plenty of promotional content, and sometimes also halluzinations/fiction/original research as well as incorrect content with malicious intent, e.g. for libel. Needless to say that this needs to be kept out of Wikidata, and this is currently done by maintaining (some level of) verifiability for each item. —MisterSynergy (talk) 00:48, 26 March 2026 (UTC)reply
I think you are setting the bar much too high. All the problems you mention are valid, but here there is a remote site that contains information about putatively-notable entities that only happen to be missing supposedly-unique identifiers. There are a lot of entities in Wikidata that do not have supposedly-unique identifiers. There are a lot of identifier schemes that are supposed to be unique but are not. So, even ignoring potential marginalization effects, not having a supposedly-unique identifier should not be a bar to inclusion in Wikidata. Peter F. Patel-Schneider (talk) 01:53, 26 March 2026 (UTC)reply
I was not referring to the example by User:Pharos above; I was addressing User:Spinster's comment directly.
We do not expect external identifiers to be present, a simple weblink to a serious source or a reference to serious printed media is easily sufficient. In fact, a Wikipedia sitelink is as well (as the implicit expectation is that there are serious sources provided in that Wikipedia article).
In terms of marginalized knowledge, there is the scenario that no references are possible, e.g. because this knowledge is not codified as we are used to in the western world into any citable (in the classical sense) sources. Verifiability is a challenge then. We need to figure out ways to include this knowledge, without opening the foodgates for all kinds of unwanted content. —MisterSynergy (talk) 18:57, 27 March 2026 (UTC)reply
What about an expansion of criterion 2 to explicitly include subjects with a record in a serious academic or GLAM database/catalog, regardless of whether there is an identifier? Pharos (talk) 01:50, 31 March 2026 (UTC)reply
@Lydia Pintscher (WMDE): Would you like some modified language specifically for criterion 2 along these lines, i.e., would this be something that WMDE would be open to considering? It would be an expansion of notability criteria for marginalized knowledge but one that is limited in scope, as serious academic or GLAM databases/catalogs are rather limited. I'm not sure where to post new language like this, perhaps we could set up a new subsection for specific proposals. Pharos (talk) 18:12, 12 April 2026 (UTC)reply
Yes, please! Let's get all suggestions on the table. LydiaPintscher (talk) 23:40, 16 April 2026 (UTC)reply
Maintainability is not solely about gaining more maintainers, but more about the minimal content of an item. I've been thinking about this, and have drafted an essay explaining my thoughts in more detail. Bovlb (talk) 16:46, 20 May 2026 (UTC)reply
  • I would very much like to hear the thoughts of editors from any marginalised communities. I agree with MisterSynergy's comments that we need to figure out ways to include this knowledge without opening the floodgates for all kinds of unwanted content. But I am firmly of the belief that this "we" should include the marginalised communities that wish to contribute knowledge to Wikidata. The suggestion "Put in place a small group of experts admins can consult for advice on matters of marginalized knowledge" to me implies that marginalised communities will be required to undertake extra steps to add their data to Wikidata. I'm assuming this is meant to apply only to marginalised knowledge that doesn't meet the proposed notability criteria. I worry this suggestion will result in not just the apppearance of but actual "gate keeping" reinforcing the marginalisation we are aiming to overcome. I urge folk to read this insightful Diff blog to gain insight into at least some of the challenges surrounding Wiki and indigenous knowledge. As a white/Pākehā woman I don't feel I have the standing nor do I have the expertise to come up with a solution to this issue without consultation with and input from the very communities it will be affecting.- Ambrosia10 (talk) 18:13, 31 March 2026 (UTC)reply
    +1 @Ambrosia10 well articulated Zblace (talk) 08:10, 19 April 2026 (UTC)reply
    +1 @Ambrosia10 Lajmmoore (talk) 18:28, 20 April 2026 (UTC)reply
    @Ambrosia10: Thanks. What my suggestion was meant to do was to address one of @MisterSynergy's earlier very valid points that as an admin they might not necessarily have the necessary skills and knowledge to decide some cases properly. And for those cases they would have the option of consulting a group of people with the necessary knowledge, skills, (lived) experience to help them make a decision. Lydia Pintscher (WMDE) (talk) 16:05, 8 May 2026 (UTC)reply
    While I agree with the spirit of what @Ambrosia10 is proposing, I insist on a point I made earlier: oftentimes, marginalized communities don't have the time or the energy to engage in long-form text conversations (and may or may not read/write in English), while they can still be very good, active contributors to Wikidata, given the nature of the project. Even for me it's a pain to participate in these conversations. I've proposed several solutions (running a banner so people can participate, conduct a few workshops, etc.), but none seem to have taken. Overall, the way in which this discussion is being conducted won't take into consideration broader community needs. Also, I couldn't agree more with @Spinster that the problem here is treating very much solvable tech problems with community exclusion dynamics. It shouldn't work like that. cheers, Scann (talk) 12:48, 14 May 2026 (UTC)reply

Input for supporting admins in decisions around marginalized knowledge

[edit]

The current draft has a carve-out for marginalized knowledge. Admins have asked for support on this matter when making notability decisions. Please add your ideas for what this could look like here.

Yes - this could be one of options, but ideally not the only option. We had admin/oversight role given to people who are mono-lingual, culturally under-informed, very young, on a spectrum, overworked (maybe frustrated with that) and inexperienced in 'real world' social issues just because of their technical skills and commitments to volunteering... so IMHO level of bad judgments this can only be secondary option and used only in high urgency when a huge volume of data from single source is 'flooding' Wikidata and even then ideally not something that gets done by a single person making a decision (or two with similar profiles and shared level of understanding). First guiding question could be: Do you have to make this decision now, fast, alone and is data likely to harm the project before you can consult relevant communities/experts? Zblace (talk) 07:24, 19 April 2026 (UTC)reply
Yes - this could be one of options, though I doubt many people wold volunteer for this kind of work and their diversity would be questionable considering that quantity/quality of volunteers varies a lot depending on economic resources and distribution of wealth across geographies/economies/genders/ethnic/language/xxx-backgrounds. So I can imagine this working better in huge language communities and much worse if at all in niches where we already failed doing more urgent and obvious work...so maybe incremental roll out of this from bigger to medium as test? --Zblace (talk) 07:29, 19 April 2026 (UTC)reply

Proposal for new criterion based on archival records

[edit]

I would propose the following, as an additional subcriterion under clearly identified conceptual or material entity:

At least one archival record: The subject is associated with an entry in a GLAM catalog or other scholarly reference work. This can be demonstrated by the use of appropriate property statements such as directly connect the subject to a distinct entry in archives, oral histories, collections, or exhibitions.--Pharos (talk) 23:12, 18 April 2026 (UTC)reply

Try to understand your change, like this fork is already notable by our criterion. As any object who is part of a museum collection. Fralambert (talk) 01:05, 19 April 2026 (UTC)reply
WD cannot be the sum of all the world's cultural databases and catalogues, recording every fragment for every archeological dig. We would be overwhelmed. Even the laudable goal of have a record of every design of manufactured object might require 100 million entries, doubling our database. We might be able to cope with the 10 million paintings in museums, but not the 5000 individual items in the Staffordshire Hoard Vicarage (talk) 06:03, 19 April 2026 (UTC)reply
@Vicarage in theory I would agree, but in practice we already have so many data of so many fields hyper-represented, that over-representation material cultural data is the least of our concerns...maybe it would be good to have the luxury of so many entries that we can off-load this in separate Wikibase for GLAMs if ever remotely close to that situation - no? Zblace (talk) 08:04, 19 April 2026 (UTC)reply
I've always thought that WD needs to be comprehensive more than it needs to be deep, hence the notability discussion. A GLAM wikibase would be a good idea, but not convinced it should be staged through WD. Vicarage (talk) 08:10, 19 April 2026 (UTC)reply
Noone is advocating for "every fragment from every archeological dig". What many of us are advocating for is the inclusion of (often underrepresented) humans who have entries in archives and oral histories. See the recent rapid grant for Wikidata-ArchivesSpace GLAM plugin. Pharos (talk) 17:39, 22 April 2026 (UTC)reply
This kind of thing is notable under the current criterion, but it would not be under the new more restrictive policy that is proposed in this RfC. Which is why I have proposed this amendment, so these would still be included. From a marginalized knowledge perspective, I am actually more interested in the lives of underrepresented people documented in archives and oral histories as in my Centro de Estudios Puertorriqueños (Q17515347) example above, rather than a standard manufactured object that just happens to be in a museum collection. Pharos (talk) 02:02, 20 April 2026 (UTC)reply
@Pharos Could you help me understand here why this would not be covered under the new proposed criteria? Is it because these artifacts don't have external identifiers?
Everyone else who has expressed concerns about marginalized knowledge: Would this addition by Pharos address your concerns? Lydia Pintscher (WMDE) (talk) 10:35, 15 May 2026 (UTC)reply
  • Support This is essential for institutional partnerships with universities, museums, libraries, government agencies, and nonprofit organizations which produce knowledge. The Wikimedia Movement needs these partnerships.
Our notability policy does not have to be a few short words which describe all possible editorial scenarios. We need a policy which helps us include millions of cultural and knowledge records, including things interested humans might want to know, but which exclude billions of research records, which are the things that only specialists or analytical bots would want to know.
For example, archeological sites dig up clay pots, but they also dig up tiny fragments of clay pots. Wikidata is a place for the pots which typical humans can comprehend, but we can exclude the fragments which are only comprehensible as a dataset for research analysis. The line is including in Wikidata things that humans would actually ponder as individual concepts.
Separating pots from fragments is not going to cleanly translate to a short Wikidata notability criteria. Importing everything with an institution's identifier is not going to work, because institutions assign identifiers to pots and fragments alike. Also, plenty of reputable institutions do not have identifiers at all, so to require identifiers as a strategy for verification is not viable either.
The point is that Wikidata should bend to match the recordmaking practices of authoritative knowledge institutions, while also keeping Wikidata as a place for humans. We should not firmly design Wikidata policies then exclude the contributions of institutional partners for failure to conform to the systems we newly create. Bluerasberry (talk) 16:30, 23 April 2026 (UTC)reply
Thanks @Bluerasberry! I think your comment is very helpful, because it brings to the discussion the scope of Wikidata and its different knowledge domains. Your example about pots vs, clay fragments makes a lot of sense. Maybe it is not the function of the notability policy to provide an answer applicable to every knowledge domain. There should be room for specific Wikiprojects to define, limit and refine the policy in terms of their specific scope, bringing more nuances and clarity to respond to questions about how far to go in terms of knowledge levels or layers within each knowledge domain. Mariana Fossatti (WK?) (talk) 21:46, 20 May 2026 (UTC)reply
@Mariana Fossatti (WK?): I wish WikiProjects were robust here, but they are not. There are about 5 active WikiProjects here, after 15 years of organizing. Wikidata has not established a workable community process for group organizing or convening discussions in the way that Wikipedias tend to do when they have this many active users. I would very much like for WikiProjects to get established here, and I wish that could have happened a generation ago because I am not sure how we can progress culturally without having a few years of organized crowdsourced discussions by region and by theme.
Right now I feel that we are compelled to have discussion at the level of global notability policy because that is the only way we can recruit enough people to have a conversation. Bluerasberry (talk) 23:38, 20 May 2026 (UTC)reply
Hello @Pharos, thanks for this proposal, it is very helpful to discuss around it. Let me know if I'm understanding well: do you mean that, in the absence of a unique identifier, an archival record from a GLAM institution can serve the purpose of comply with notability criteria #2. In sum, if you have the ID, or a serious source or a GLAM record, you are fulfilling the criteria, correct? Is this a third aspect of N2, or it is a complementary condition or alternative condition in regard with identifiers? Asking this just to get more clarity on how to implement the proposal, but I agree with the spirit of the idea. Many important entities don't have external identifiers, but they can be clearly identified with other methods, and GLAM institutions can provide proper identification even when they don't always provide an ID or the ID isn't yet adopted by Wikidata. Mariana Fossatti (WK?) (talk) 21:35, 20 May 2026 (UTC)reply
Yes, my intention was that you'd would fulfill that criteria by having either the ID, or a serious source or a GLAM record. I guess I would see it as a third aspect of N2, but it could also I suppose function like a complementary condition.
Have also been considering the development of an outreach:Archive of Archives for under-resourced GLAM efforts, that may benefit from an on-wiki host for some of their records. Pharos (talk) 03:00, 25 May 2026 (UTC)reply
Thanks for the clarification, it makes sense! The Archive of Archives project looks interesting, I'll have a look. Mariana Fossatti (WK?) (talk) 15:33, 25 May 2026 (UTC)reply
  •  Oppose To me, this proposal is incomprehensible and must therefore be rejected in its current form. Why should a GLAM catalog require its own separate entry, rather than falling under the category of a "reliable source"? --Gymnicus (talk) 20:42, 25 May 2026 (UTC)reply
    This proposal was formulated as a response to the CENTRO question above. While I'd like to think that GLAM records are a reliable source per se, apparently not everyone agrees with that definition. And certainly many GLAM records do not fulfill the criteria of being an "independent" reliable source, which is another part of the criteria this would be supplementing. Pharos (talk) 18:48, 27 May 2026 (UTC)reply
    I still don't understand the point. For when I read through the CENTRO discussion above, the issue there isn't the independence of the source, but rather verifiability.
    For you state that the data researched there are not published via identifiers. How, then, are they published? Are they published only in an internal database that is accessible from the outside? --Gymnicus (talk) 09:02, 30 May 2026 (UTC)reply
    A lot of GLAM data are published but not in a structured way: for example, in a public website with pages for each object, or in a table in a document, or in a printed catalogue. The GLAM institutions that have database-driven web sites are a minority and they are usually in the richest countries. The distinction between GLAM catalogues with database-like structure versus unstructured is not very relevant to Wikidata's interest. There are nicely structured archaeological databases with hundreds of thousands of things dug out of the ground, almost all of which are non-notable for WD. At the same time, there is heritage described only in Word documents posted on a website that is nonetheless recognised by scholars as the most important in the world, where making a document is all the technology that under-resourced museum or archive can manage. MartinPoulter (talk) 16:43, 1 June 2026 (UTC)reply
    Such a document can and should be added as a source. Bovlb (talk) 16:45, 1 June 2026 (UTC)reply
    I have have made a related proposal for a representation of GLAM topics on Outreach-Wiki, trying to give a Wikimedia home for some of these unstructured documents from less-resourced institutions. Pharos (talk) 15:09, 2 June 2026 (UTC)reply

Multi-sensory knowledges

[edit]

I suspect this isn't quite the right place, but I just wanted to flag multi-sensory data as an area that Wikidata in general is very limited, but there is huge potential. However, despite taste, smell and touch being innately human, there's very little listed here. To give an example - smell-related data has expanded over the past two years, but it is dominated by perfumery - because there are many external catalogues for perfumes, but not for more everyday smells (even though we know they exist). Similarly for taste; haptic data is the most scarce. Whilst I understand there are needs for Wikidata to alter, I would really encourage thought to be given to other types of knowledge that we would wish to include (to reflect the world) in the future.

Round 2: External identifiers: When do they support notability and when do they not

[edit]

As pointed out in the discussion a core part of the new policy are external identifiers. Wikidata has over 10.000 of them by now. Some of them should support the notability of an Item and some should not. We currently do not have explicit criteria or processes for how to decide when a Property should be instance of (P31)Wikidata property for an identifier that suggests notability (Q62589316) or instance of (P31)Wikidata property for an identifier that does not imply notability (Q62589320). This would be a requirement though for putting a new Notability policy into practice. Based on which criteria should we classify a Property into one or the other group?

properties which should always have instance of (P31)Wikidata property for an identifier that does not imply notability (Q62589320):
I am honestly unsure how to deal with hugely-used IDs like ISBN-13 (P212), ISBN-10 (P957), DOI (P356) but surely saying that they always imply notability would be very dangerous; however, saying that they do not imply notability could lead to the mass-deletions of millions of items ... as of now I don't have a good solution in mind. --Epìdosis 11:07, 22 March 2026 (UTC)reply
For example, property GND ID falls under Wikidata property related to a thesaurus. A data object with this property has been considered notable for years. I certainly use this myself. But nevertheless, you must ask: What rules exist within the Integrated Authority File and also within other authority records for creating an identifier? Because this isn't entirely clear for all records. --Gymnicus (talk) 13:58, 22 March 2026 (UTC)reply
we need to remember wikipedia is a community project, and sites like IMDB get more oversight than a small wikipedia. Often items that appear in a Wikipedia list can be fully fleshed out on a community site. Do we have a property that connects WD items to WP lists. While buildings often have national registers, objects and creative works don't. As for ISBNs given even minor publishers get issued large blocks, their presence is no guarantee of much. I'd work on the principle that external IDs give notability unless explicitly recorded otherwise Vicarage (talk) 13:09, 22 March 2026 (UTC)reply
@Vicarage IMDB don't imply notability is not so bad, since a person in it could be a structural need for a movie. As for book, my cut for notability is if a work as been self-plublished or not. Publisher will often create a ID in their national library, making their author notable.
That kind of exclude most of self-publisher work, but my system is not perfect, as it will imply that all article in a newspaper are notable, which I don't agree. Fralambert (talk) 23:06, 22 March 2026 (UTC)reply
Declaring traditionally (not self-published) works to be notable seems like an eminently reasonable policy for books, as such books presumably have to make it through some form of editorial review; at the very least I'd consider anything more than a few decades old (say, anything pre-2000) to be notable, as such books are highly unlikely to be self-promotional (most self-promotion is done by self-published writers to begin with). A full record of books, at least traditionally published ones, which can be searched by such things as publisher, publication date, genre, etc., is something which, to my knowledge, does not exist (Worldcat comes close, but doesn't quite meet my needs, as it's focused specifically on library records, and searching it isn't as easy as querying Wikidata; ISFDB comes closer but only covers certain genres; others are usually either missing too many books to be useful for my purposes or have poor search functions); I've long wanted a site where I can search something like "books published in paperback in 1977" or "all books published by [insert publisher here]" or "every mystery novel from 1943" or the like, and Wikidata would be ideally suited for this purpose. I normally stay out of policy discussions but I feel strongly enough on this issue to post here, seeing that I mostly edit in the field of books, especially older ones; if we exclude books that are not held in national libraries, for example, we then lose such use cases as finding how many books a specific publisher published in a given year or comparing how many books were published in a given genre over the course of a span of years since the data will be incomplete, and thus inaccurate.
I'll also mention that ISBNs weren't universally in use until sometime in the 70s and many libraries didn't stock paperbacks during the early decades of that format's existence, so some notable books from the 50s, 60s, and early 70s may not have associated ISBNs or library identifiers. TenebrousFox (talk) 08:32, 23 March 2026 (UTC)reply
Internet Speculative Fiction Database (Q2629164) is an interesting case. Its clearly a community resource, and can be astonishingly comprehensive. A search for "Journey Planet (Q105759665)", which a self-published SF fanzine, notable in its field, and has en:WP and WD entries, returns 598 records, as someone has lovingly recorded every article, and every piece of art work. Yet "Attitude (Q115888549)", another important fanzine from 20 years ago, has no entries in ISDFB, and a WD entry because I thought it worthy. ISFDB has multiple identifiers here , but they do not cover all its categories. Fancyclopedia (Q42964695) has an external identifier here, and covers both fanzines, but also refers to the most obscure single page sheets distributed to 100 people. I've no idea how we'd provide consistency of coverage if we apply rigid rules on identifiers, or respond someone argues that a fanzine book review's ISFDB presence equates to an academic journal DOI for a paper. I've been adding fanzines to WD, but my criteria is rather arbitrary, they should have scanned copies online, so the entry has some body.
In lots of fields a WD enthusiast can push the boundaries a long way from the norm, but others, less enthusiastic, can force deletion. Fundamentally, notability is hugely variable in the outside world, and is likely to be so in WD. Individual notability can always be challenged, perhaps it should be considered for a body of work, like 1000 SF fanzines, being useful as a whole and so notable. Vicarage (talk) 11:49, 23 March 2026 (UTC)reply
Are we certain that having an ISBN implies notability? There are tons of self-published ebooks with ISBNs; declaring them all notable simply by virtue of having an ISBN opens the project up to a lot of self-promotional nonsense. Omphalographer (talk) 01:44, 2 May 2026 (UTC)reply
It's at least verifiable that a given ISBN can be connected with a real-life print, audio, or ebook. If the item isn't notable the ISBN doesn't make it notable, but is one more way to determine that the publication in question isn't a total fiction. Arlo Barnes (talk) 19:20, 17 June 2026 (UTC)reply

I have asked our data analyst to check the current state of coverage for instance of (P31)Wikidata property for an identifier that suggests notability (Q62589316) and instance of (P31)Wikidata property for an identifier that does not imply notability (Q62589320). The analysis is available in this Google spreadsheet and has some additional info like how many Items the Property is used on etc. You can make a copy and play around with it if you want. I have also created basic Listeria reports at Wikidata:Notability/External identifiers for all 3 categories. What this tells me is that we have a ton of Properties that are not classified and if we want better automated tooling around determining notability we need to get the coverage up. What is blocking that currently? Lydia Pintscher (WMDE) (talk) 11:21, 31 July 2026 (UTC)reply

@Lydia Pintscher (WMDE): thanks very much for the Google spreadsheet, this is very useful. I 101% agree on the conclusion "we have a ton of Properties that are not classified and if we want better automated tooling around determining notability we need to get the coverage up". In my opinion, the reason why the coverage is so scarse is that we effectively miss any list of criteria about how to apply this two markers (and also what to do in case of disagreements about the already applied ones ... I do not remember as of now disputes on this topic, but they might of course happen in the future), apart from common sense. And common sense might be enough to solve a fair percentage of cases, but I think that in at least half of them a small set of criteria is inevitably necessary. Just as an example, out of the first five external IDs by number of uses Google Knowledge Graph ID (P2671) (7.9 M uses) reasonably does not imply notability, SIMBAD ID (P3083) (8.1 M uses, 3.27% with sitelinks) implies notability (note: potentially at least 9 M other values could be imported), PubMed publication ID (P698) (32.1 M uses, 0.06%) DOI (P356) (28.9 M uses, 0.23%) PMC publication ID (P932) (6.6 M uses, 0.09%) are unspecified ... and I am unsure if there is consensus about the choice to be made. Epìdosis 12:56, 31 July 2026 (UTC)reply
@Lydia Pintscher (WMDE): User:Epìdosis/highlight identifiers notability.js just created (with the help of Claude); it clearly shows in an item which IDs imply notability (green) and which not (red). Epìdosis 08:04, 1 August 2026 (UTC) P.S. I have cleaned up Wikidata:WikiProject Properties/Participants so it can be used as {{Ping project}} (until 7+ new users subscribe to it, cf. phab:T200350 + phab:T148154). --Epìdosis 08:56, 1 August 2026 (UTC)reply
I have used my gadget to discover some tens of relatively easy cases and mark them; so now the numbers are a bit down, I put here a query for reference (https://w.wiki/T3ud); as of now, out of 10454 external IDs, the results are 660 properties do not imply notability (6%), 1742 properties suggest notability (17%) and 8052 properties are not classified (77%). Epìdosis 17:06, 2 August 2026 (UTC)reply
Currently, adding these is a matter of individual user decisions rather than community decisions. That is why I am unsure how meaningful this gadget and statistics actually is. For example, the property Davis Cup player ID until 2025 (archived) is marked as Wikidata property for an identifier that suggests notability. This is understandable. After all, in most versions of Wikipedia, national team players are considered notable, and every Davis Cup player is a national team player. On the other hand, the property World Athletics athlete ID is marked as Wikidata property for an identifier that suggests notability, too. Many of the athletes listed there are certainly notable for Wikidata, but by no means all of them. Consequently, this criterion must either be removed from the list, or the links must still be individually verified in every case.
I actually touched on this earlier myself, that I see issues with authority data. For outsiders like me, it isn't clear why one entity receives an identifier while another doesn't. Take the sports club SC Tegeler Forst, for example. It has a GND identifier. In contrast, the club Q15431372 does not. On what basis is this decided? And the far more important question, how do we handle such ambiguous situations? --Gymnicus (talk) 11:50, 9 August 2026 (UTC)reply
  • I open a new point for a related question: can obsolete IDs (cf. phab:T306761) imply notability? If an ID implies notability and then it becomes unavailable (website shut down, not archived or only partially archived in the Wayback Machine) what do we do with items based solely on this source? I don't have a good answer as of now, but I think this issue is relevant enough to be discussed here. --Epìdosis 21:43, 11 August 2026 (UTC)reply
    I think yes - the Wikidata entry might be more valuable as a record of what the originating site said, even if it is no longer available. That said it leaves the data less verifiable so it's not a good long-term situation, hopefully some other identifier becomes available as a replacement. ArthurPSmith (talk) 19:20, 12 August 2026 (UTC)reply

Round 3: Context

[edit]

Below is the revised proposal for the new policy based on the feedback on the first proposal. Please add your votes, questions and comments below.

Before reading the new proposal, I encourage everyone to read User:Bovlb/maintainability. Lydia Pintscher (WMDE) (talk) 07:38, 9 June 2026 (UTC)reply

Changes from the first proposal:

  • Added Wiktionary to criteria 1 project list based on request.
  • Added a sentence to criteria 1 to indicate that the presence of a Wikisource index page URL (P1957) statement can make an Item notable based on the request of Wikisourcerers.
  • Tweaked criteria 2.1 to add archival records based on the suggestion by Pharos and resulting discussion. I didn’t make it a separate point entirely as it is very close to the existing identifiers point.
  • Revised sentence about conflict of interest editing in order to focus on prohibiting self-promotion and commercial marketing editing. Those seem to be the most egregious offenders, especially recently, creating a lot of work for admins while clearly not contributing in the interest of Wikidata.

Round 3: Policy proposal

[edit]

Wikidata is a knowledge graph, providing general purpose data to the Wikimedia projects and the world at large. Wikidata is intended to provide data for a basic understanding of the world for humans and machines.

This policy defines which entities can be described in Wikidata.

An item is acceptable if and only if it meets at least one of the three criteria below:

  1. It contains at least one valid sitelink to a page on Wikipedia, Wikivoyage, Wikisource, Wikiquote, Wikinews, Wikibooks, Wikidata, Wikispecies, Wikiversity, Wikimedia Commons or Wiktionary.
    1. <most of subpoints of criteria 1 remain unchanged except for subpoint point 5>
    2. 5: On Wikisource, items for main namespace pages, Author pages, Translation pages, and Portal pages are valid, along with items for namespaces that exist on other Wikimedia sites (Category, Project...). Pages in the Index and Page namespaces are not considered valid. Subpages of main namespace pages (for example, individual chapters) are not considered valid unless they are written by separate authors, could be considered stand-alone works, or have been translated separately. Additionally the presence of a Wikisource index page URL (P1957) statement indicates notability.
  2. It refers to an instance of a clearly identified conceptual or material entity that is described using serious and publicly available references. The minimum criteria:
    1. At least one external identifier or archival record:
      1. The subject is associated with one or more identifiers that are assigned or curated by an independent authority, and that meaningfully distinguish the subject from others. Identifiers that are self-assigned, or can be obtained without editorial review (such as typical social-media profiles) are not sufficient on their own.
      2. The subject is associated with an entry in a GLAM catalog or other scholarly reference work. This can be demonstrated by the use of appropriate statements such as directly connecting the subject to a distinct entry in archives, oral histories, collections, or exhibitions. OR:
    2. At least one reference, independent from the subject: The subject is the topic of significant, independent coverage in serious and publicly available sources that support a meaningful identification and distinction from others. Routine listings, directory entries or being quoted in an interview about another topic do not constitute significant coverage on their own, and additional independent sources might be required. Self-authored descriptions, and marketing or promotional material are not considered serious and independent sources.
  3. It fulfills a structural need, for example: it is needed to make statements made in other items more useful.

New items must have the statements necessary to be clearly identified or sitelinks that show that they meet at least one of the above criteria within 24 hours.

In order to avoid the appearance of a conflict of interest, creating an Item about yourself, your close relatives or your business is not permitted. Additionally we do not allow commercial companies to create Items about their clients or their products for marketing purposes. If any of them meet Wikidata’s notability criteria, someone else might create them for you eventually.

Please consider

[edit]

Wikidata’s data is maintained by a community of people like you. Please be considerate of their time and effort. This is important for the long-term health of the project. Before creating a lot of new Items:

  • Spend some time editing existing Items to get an understanding of data modelling and best practices first. We value quality over quantity.
  • Think about where and how your data will be used. If your data is used in a Wikimedia project or another application, chances are higher that it will be kept in good shape long-term and Wikidata’s overall data quality stays high.
  • Consider what the complete set of Items would be that you are working on. Wikidata is not the right place to describe all stars in the universe, every human on earth, etc. Narrow it down to a reasonable set of notable entities and discuss with the community in case of any doubt.
  • If you intend to create a lot of new Items and are not absolutely sure if they meet the notability criteria, please ask for additional input at Wikidata:Notability/Questions.
  • Do not try to game the system. Such efforts are frowned upon and will only lead to your Items being deleted.

Knowledge equity

[edit]

We do acknowledge that certain knowledge has been and is being structurally marginalized. This leads to less coverage in reliable sources and therefore an increased barrier to demonstrating notability. In our effort to make more knowledge accessible to everyone, we take this limitation into consideration when making decisions about the notability of an entity.

Options besides Wikidata

[edit]

If your data does not fit Wikidata’s notability criteria, we want to help you find the perfect home for your contributions. Wikibase Cloud and Wikibase Suite might be a better fit for your data. Together with Wikidata, these form the Wikibase Ecosystem, a network of specialized communities, each collecting and maintaining data for the world in their area of focus.

Data in any Wikibase instance becomes part of the Linked Open Data Web and accessible worldwide. In addition you have more flexibility in how to model your data and who gets to edit it. Visit wikiba.se to explore which platform best suits your needs, and check out our showcase to see examples of what other communities have built.

Other entity types

[edit]

This notability policy covers Items. The notability criteria for Lexemes can be found at Wikidata:Lexicographical data/Notability.

Round 3: Collecting concrete examples

[edit]

In order to help everyone better understand the concrete impact of the proposed new policy, let’s collect examples here.

Currently notable but would no longer be notable:

  • Q50689460 a dolmen, but was created empty with no sources.
  • Q141700881, a self-published book (+ author and editions), created by its author probably using an LLM-based tool. It could be considered notable under the current criterion 2 because it has an ISBN, publisher, and publication date. Under the proposed wording, it would likely not be considered notable because its identifiers (self-published ISBNs, ORCIDs, and self-deposited DOIs) can be obtained without meaningful editorial review.

Currently not notable but would become notable:

Round 3: Discussion

[edit]
  • I moved the examples to a separate section as it would have gotten mixed up with the wider discussion (feel free to undo). Regarding the Q50689460 example, somebody has added an external ID since. Independent from whether this makes it notable or not, it raises questions about what happens to items that are already in Wikidata but lack notability according to the new policy. Will they be batch deleted? What if there are items that could be easily given external IDs (for example through Mix'n'Match) but for some reason it hasn't happened yet? Would there be a grace period for non-notable items before deletion? Also, in case we go with this policy, users creating new items without external IDs should be given clear guidance about the possibility of their item being removed. Adam Harangozó (LSE WiR) (talk) 17:54, 10 June 2026 (UTC)reply
    Also, I think it would be important to deploy a policy like this with care for research data, meaning not to scare away researchers and institutions who come here to share open data (and whom we are otherwise trying to lure in). Ideally this would mean that there is easily findable and detailed information for adding research data or metadata, helping them decide when to choose Wikidata and when Wikibase. But this would only make sense if Wikibase would have lower entry barriers, for example researchers wouldn't need to create an ontology from scratch but would at least have the option to easily copy a basic structure from Wikidata. Otherwise what this achieves is deterring researchers from using both Wikidata and Wikibase. Adam Harangozó (LSE WiR) (talk) 18:03, 10 June 2026 (UTC)reply
    @Adam Harangozó (LSE WiR) I added this exemple because it was proposed in RfD yesterday and the vestion as at this stage when propose. While a dolmen (Q101659) could generally fulfil notability, this item didn't have source at all when propose. There is also hown we should treat a RfD because someone dumped a database in 2016 with mostly no source. Fralambert (talk) 21:49, 10 June 2026 (UTC)reply
  • Many thanks for putting together a 3rd draft of the policy proposal, Lydia!
I'd appreciate clarification on what we (the Wikidata community) mean by archival record. For example, I know that all of the information in this database is based on archival records because I worked on this project for four years in my previous job. Very few of these records are online, but as we've discussed in round 2 that isn't necessarily a problem since most archives cannot afford to digitize all of their material, and we want to be inclusive of marginalized groups and their histories. The way I read point 2.1.2 in the 3rd round policy proposal suggests that all 2,491 people in this database are notable. Many of these people would not be considered notable under the existing policy because they can't "be described using serious and publicly available references" (emphasis on plural is mine), and don't make other Wikidata items more useful. Under the revised policy, the word "references" is followed by "The minimum criteria is: At least one external identifier or archival record". In this case, each of these entries is associated with a "distinct entry in an archive," even though most of the entries are for ordinary people and their families who the FBI decided to arrest during World War II.
As a more specific example, I had a really hard time finding more sources for Ichida Heikichi (市田平吉), a fisherman, aside from what's already in the database. All I could find is a death notice from 1962 without enough context to confirm that this is the same person, especially since the last archival documentation of him states that he was repatriated to Japan after the war. Are we okay with having these kinds of difficult-to-verify items in Wikidata?
To be clear, I am not asking this because I want to import everything from this database into Wikidata. I don't work for that organization anymore, so it's not my decision to make. I am asking this question so we can be sure that this is what the new policy should mean. If we decide that's exactly how this policy is to be interpreted, great! But if it's not, I think we should be aware of this reading of the policy and adapt accordingly, perhaps by changing the OR at the end of 2.1.2 to AND. Just a thought, Mcampany (talk) 21:03, 10 June 2026 (UTC)reply
I can move the example up to the "Currently not notable but would become notable" section if desired, I just wasn't sure if it would make sense for it to be there without the additional context I've provided. Mcampany (talk) 21:16, 10 June 2026 (UTC)reply
@Mcampany: Thank you for pointing this out. To make sure I understand correctly: You're basically saying that in the shuffle we moved from requiring several references to requiring only one? Good catch. That was not my intention, but then I'm also not sure we're actually requiring several right now...
What do others think? Lydia Pintscher (WMDE) (talk) 09:41, 31 July 2026 (UTC)reply
I don't think we are presently requiring more than one (good) reference, and in my opinion one (good) reference is enough to create an item. Epìdosis 12:59, 31 July 2026 (UTC)reply
Yes, exactly. The current notability policy reads "serious and publicly available references", and the presence of an "s" at the end of "references" suggests (to me) that an item needs more than one source to be notable. Whether this actually happens in current practice is debatable, but I think that using the phrase "at least one" in the new notability policy suggests that only one external identifier, archival record, or independent reference is required. I have no particular opinions about this, but I wanted to flag that difference since it seems like a major change. Mcampany (talk) 13:37, 31 July 2026 (UTC)reply
I generally consider that one good reference is sufficient to pass the source criterion, notwithstanding the "s". Best practice is to have multiple independent references, of course. There's sometimes a tension between evaluating the bare minimum requirement and encouraging people to follow best practice, and I get feedback that these are not always distinguished well. Bovlb (talk) 15:39, 31 July 2026 (UTC)reply
  •  Comment Lydia, I appreciate the time and effort you're putting into this. However, I think that there are several things that still need to be improved:
On the "serious" criteria: the section on "At least one reference, independent from the subject" could benefit from a more neutral language. What constitutes "serious" publications can vary among geographies and cultural contexts, so it would be important to look for more objective ways to establish the particular quality that you're referring to.
On the quotation criteria: "Being quoted in an interview about another topic do not constitute significant coverage on their own" is an unclear sentence and hard to understand by a non-native English speaker (such as myself). An environmental defender will very unlikely be directly interviewed about themselves, but rather they'll always be interviewed about the topic that they work on -- the sentence should read: "Being quoted in an interview about a topic unrelated to the area of work or activism of the person itself".
On the "additional independent sources": for cases such as the one I'm describing (an env. defender) you'll have ten interviews that might quote more or less the same, independently of the relevance or importance of the person and the news outlet publishing the interview. I'd suggest also establishing a min./max. of sources to be provided to determine the relevance, otherwise you'll end up with five links that say more or less the same about a person.
On the sources list: the sources list is very skewed to traditional, Western media, such as scholarly reference work, GLAM institutions, and other "serious and publicly available sources" (i.e., the press). Consider adding an exception for marginalized knowledge that clearly allows for specialized media in any format considered to be a significant authority in the matter (i.e., YouTube channels).
On the acknowledgement of marginalized knowledge: this is the most problematic statement of all of them. Who is the "we" talking here? Who takes the limitation into consideration? And what's the criteria? This needs way more work; the limitations & exceptions criteria needs to be crystal clear, the procedure has to be clear, it has to be clear who is the "we" that takes that limitation into consideration, and it needs to add clear procedures to fix and repair the issue if marginalized knowledge is erased without good justification. Just saying that we acknowledge something and not offer any avenues for repair is not enough.
Consider adding a section on revision of the policy after it has been implemented for six months. I've made this comment back in the day and it was ignored, but I think it's important to actually evaluate the impacts of the policy to finally establish it.
cheers,Scann (talk) 15:45, 16 June 2026 (UTC)reply
@Scann: Thank you.
Do you (or anyone else) have concrete wording suggestions? I'm generally happy to adapt but stuck on how to improve this.
As for your last point: @Mariana Fossatti (WK?) has been working on this and will hopefully share it here soon. Lydia Pintscher (WMDE) (talk) 10:01, 31 July 2026 (UTC)reply
@Lydia Pintscher (WMDE): Probably it is User:Mariana Fossatti (WK?)/Wikidata Knowledge Equity Guidelines published on July 19. Epìdosis 13:00, 31 July 2026 (UTC)reply
Thanks for the link. It's a thoughtful document, and I added some comments on it's talk page. Bovlb (talk) 04:29, 2 August 2026 (UTC)reply
Indeed, I've moved the page here, and also I've added a new section in the discussion for make it easier to find it in this discussion: Wikidata:Requests for comment/Notability policy reform#Knowledge equity guidelines. Thanks @Lydia Pintscher (WMDE), @Scann, @Epìdosis and @Bovlb for your interest and comments! Mariana Fossatti (WK?) (talk) 21:41, 17 August 2026 (UTC)reply
  •  Comment The sitelink criteria is too strict for proper interaction with Commons. For many items, such as artworks / artefacts items, there might be only one picture. Take for instance Q134777601. I thus propose to add to criteria 1 "or the file used in image (P18) uses wikidata in its description". Léna (talk) 12:02, 6 July 2026 (UTC)reply
    @Léna: I am not sure I understand. Are you talking about an Item that is used in a statement on a file on Commons and should therefore be notable? Lydia Pintscher (WMDE) (talk) 10:04, 31 July 2026 (UTC)reply
    Yes ! For instance Q139793315. For this artwork we are lucky the museum published an exhibition catalogue, but it is not always the case and Wikidata is used to store the information on the cartel and display it on Commons. It's a shortcut from the "imported from wikimedia commons" source value. Léna (talk) 08:14, 5 August 2026 (UTC)reply
  •  Comment for this "In order to avoid the appearance of a conflict of interest, creating an Item about yourself, your close relatives or your business is not permitted. Additionally we do not allow commercial companies to create Items about their clients or their products for marketing purposes."
In English Wikipedia it's said: "The practice of editing or creating articles about yourself, your family or friends, your organization, your clients, your employers, or your competitors is strongly discouraged." So it is allowed even though discouraged. Why should Wikidata be more strict in this?
What I would suggest is: ask editors be transparent about their COI edits. Of course they cannot edit items that doesn't meet the notability criteria and if they continue to break the rules they could be banned. If the white hat editing is not permitted here there will be a lot of black hat editing from sock farms anyway so changing the policy may just generate more work for admins. I think its a bad idea to drive the people away who have the knowledge and will to improve the items. I've activated here in November and while I've edited things for my customers I've made plenty of improvements to related items, e.g. corrected (and added missing) translations. There are not that many Finnish Wikidata editors here anyway.
I also think that the active Wikidata volunteers might occasionally edit for items they have some COI with. So what happens if a volunteer with thousands of edits is get caught about editing their family member's item, say a grandparent? While the editors are mostly anonymous we might not ever know about this but the editors themselves might be feeling quilty about breaking the rules and as a result they may want to stop contributing.
Another idea: Ask the commercial companies to A) acquire a certificate that allows them to edit and create Wikidata items or B) create xxx meaningful edits from their knowledge area before starting to COI edit.
Cheers Jjanhone (talk) 18:31, 7 July 2026 (UTC)reply
So it could be like this: "In order to avoid the appearance of a conflict of interest, creating an Item about yourself, your close relatives or your business is allowed only for editors with 150 edits. In the same time they are required to disclose the name of every organization (regardless of legal status) and individual that pays them for contributing to Wikidata." (The latter part is from [2].)
See also [3]: "Conflict of Interest is broader than paid contribution: e.g. editing the item of your favorite band, about yourself, your friends or rivals, your home town, can all lead to edits with a Conflict of Interest. Such editing in Wikidata is not frowned upon in general, but can indeed be quite helpful: who else but those with a strong interest in a topic would be editing the topic?" Jjanhone (talk) 18:54, 7 July 2026 (UTC)reply
@Jjanhone: We landed on prohibited because people said earlier in the discussion that if it is just discouraged it has no meaning.
What do others think of @Jjanhone's suggestion here? Lydia Pintscher (WMDE) (talk) 10:08, 31 July 2026 (UTC)reply
I had expressed myself similarly like Jjanhone on the subject in one of the earlier sections of the RFC. He has now, of course, articulated it more clearly and I must say that I like the way it is phrased. --Gymnicus (talk) 22:06, 17 August 2026 (UTC)reply

AND vs OR in criteria 2

[edit]
  •  Comment (Repeating an earlier comment because I was the last person to respond to Round 2 before it was collapsed.)
I like the new wording for identifiers and sources. The proposed phrasing seems to bring them more closely into line with common practice at RfD and AN (requests for undeletion).
I remain concerned about the change from "AND" to "OR". This seems by far the most significant part of the proposal, and we should be clear about its implications.
For me, N2 is fundamentally about maintainability. For Wikidata to succeed in the long term, it must be possible for editors to be able to maintain items without relying on the original item creator. That requires both identity (knowing when two records should be merged or kept separate) and sources (allowing information to be verified, updated, and extended). Ensuring both of these is the reason N2 currently uses "AND" rather than "OR".
Regarding GLAM and archival records, do we have any concrete examples of entities that should be considered notable because of an archival record that would not qualify under the current policy? Bovlb (talk) 21:10, 9 June 2026 (UTC)reply
N2 currently doesn't say anything about External identifiers. The only "and" in the current single sentence is for "serious and public" but that is referring to the quality of a reference that could be added. Ainali (talk) 04:26, 10 June 2026 (UTC)reply
I’ve been reflecting on the various different ways that people understand notability. I think there is a unifying principle behind the policy: the collaboration point, the transition when the baton is passed from the creator to the community.
Before this point, accurate interpretation of the item relies on information known only to its creator. After it, the item has been sufficiently tied into external or internal resources that other editors can take over its stewardship: verifying, distinguishing, merging, correcting, and extending it.
Seen in this light, the proposed wording change can be understood as making the collaboration point more explicit. It shifts the emphasis away from an item’s potential to become collaborative and towards demonstrating that it already is. Bovlb (talk) 05:57, 4 July 2026 (UTC)reply
  •  Comment Firstly, many thanks to Lydia for the great work on this. A couple of points: It's not really clear to me whether for Criterium 2, both sub-criteria "At least one external identifier or archival record" and "At least one reference, independent from the subject" need to be fulfilled. It seems to be the case, but I think there's room for doubt or misinterpretation. I wonder if there's a way to make it a bit clearer. Also, "If your data is used in a Wikimedia project" threw me a bit at first since Wikidata itself is a Wikimedia project, but I guess it's referring to other Wikimedia projects. Again, maybe it's something worth clarifying. --Yirba (talk) 23:55, 11 June 2026 (UTC)reply
    There is an "OR:" in the text. I think this should be made more clear in the next round.. So9q (talk) 17:19, 30 June 2026 (UTC)reply
  • Sorry if I have missed it in the above discussion, but I think we need an "and" or an "or" between 2.1 and 2.2. It isn't clear to me if those are to be taken together (as in, an item to be notable must have one identifier or archival record and a reference independent from the subject). Happy to boldly reword the proposal a bit to conform with the intent once I know it. -- Ajraddatz (talk) 06:13, 23 July 2026 (UTC)reply
    Looks like the "or" is already there, it's just attached to the end of 2.1.2 as opposed to outdented and nested between 2.1 and 2.2. Yirba and So9q both mentioned this about halfway upthread. It's redundant where it is now, since 2.1 already says "[a]t least one external identifier or archival record". I read that as either 2.1 or 2.2 will suffice. Lydia Pintscher (WMDE) can you confirm that's the intent before Ajraddatz or anyone else rewords it? As for which one it should be, I think this is the better call for practical reasons: a requirement to satisfy both would bar items covered in serious sources that have no identifier or archival entry, which seems too restrictive for the project. WWB Too (talk) 16:39, 27 July 2026 (UTC)reply
  •  Question I've moved all comments about the AND vs OR into this dedicated section so we can come to a conclusion. I had put OR and it was AND in earlier unpublished drafts. I'd love to hear more input: Should it be OR? AND? Or does anyone have another different idea to work around this? Lydia Pintscher (WMDE) (talk) 10:15, 31 July 2026 (UTC)reply
    In my opinion surely OR, because requiring "At least one external identifier or archival record" AND "At least one reference" would exclude from notability a lot of entities which are described by reliable printed sources but not (yet) in online databases having an external-id property in Wikidata (from my comment of June 9). Epìdosis 13:05, 31 July 2026 (UTC)reply
    Agree with Epìdosis in favoring OR. Requiring both would exclude subjects covered by reliable sources but lacking an identifier or archival coverage, which is too restrictive. WWB Too (talk) 00:57, 13 August 2026 (UTC)reply
    I support the new proposed policy as far as it is clear that in N2 we have an OR, not an AND between identifiers and references. I think is the spirit, but I see also confusion on how to interpret the text. Mariana Fossatti (WK?) (talk) 21:15, 17 August 2026 (UTC)reply
  • My support is conditional on this being an OR, as said below. There are plenty of people who are Wikipedia notable and who have full books about them who don't have identifiers. We shouldn't be forced to not include these people until they get a Wikipedia article. PARAKANYAA (talk) 02:56, 13 August 2026 (UTC)reply
    Also, to clarify further, there are many classes of thing for which no identifier types exist. This would make creating items on them impossible regardless of how much sourcing there is. PARAKANYAA (talk) 03:23, 10 September 2026 (UTC)reply
  • I am a member of the wikiproject "Public Domain in Latin America" and the project "Strengthening the Latin-American Public Domain with Wikidata." In those communities, we create and improve items about authors in public domain, works, cultural institutions, etc. We've been discussing the proposal in depth. For us, and I think this is true also for other communities working with knowledge from the Global South, the presence of the OR between 2.1 and 2.2 in the Notability Policy is extremely necessary, given that sources are more difficult to access than in other contexts where cultural institutions are well funded. For example: - Sometimes institution identifiers exist, but other times important institutions do not provide official identifiers for their entries. - Sometimes we don't find records in cultural institutions, but rather references in old cultural magazines, newspapers, or other old publications. Cultural institutions often have enormous gaps in their catalogs. For all these reasons, I greatly value the addition of the OR between 2.1 and 2.2, and also the addition of section 2.1.2 about archival entries in GLAM catalogs that don't have identifiers. Both are positive steps, and I will support the proposal as long as they are maintained. Pepe piton (talk) 16:53, 28 August 2026 (UTC)reply

Who are we to decide what is notable for the entire world?

[edit]

There is a much more fundamental problem here that I do not think this proposal addresses at all:

Who are we to decide, from a predominantly Global North perspective, what is or is not notable for every community in the world?

Imagine a community whose knowledge is transmitted primarily through oral tradition, with little written documentation and virtually no digital footprint.

Who exactly are editors sitting in London, Berlin or elsewhere in highly digitised societies to decide that a person, organisation, cultural practice or body of knowledge within that community is "not notable"?

Not documented according to your standards does not mean not important.

Not digitised does not mean not important.

Not covered by newspapers does not mean not important.

Not catalogued by an institution that Wikidata editors happen to recognise does not mean not important.

And absence from the Internet certainly does not mean absence from human knowledge.

This is not a hypothetical concern.

We already have editors applying unwritten standards and personal essays to argue for the deletion of books and items about scientists from developing countries because they supposedly lack sufficient "notability". Now the proposed solution is to make the notability policy even stricter.

That is absurd.

The problem is not merely that some communities have fewer sources. The problem is that we are defining what counts as a valid source, what counts as sufficient documentation and ultimately what counts as a human being, book, organisation or cultural entity worthy of representation according to standards developed in societies with radically different documentary infrastructures.

A scientist from a wealthy country may leave behind university profiles, ORCID records, Scopus entries, institutional repositories, newspaper coverage and multiple national authority identifiers almost automatically.

A scientist doing equally important work in a poorer country may have a CV in a university drawer, publications in journals that were never digitised, records in databases inaccessible from abroad and almost no searchable online presence.

Then an editor sitting thousands of kilometres away looks at Google, finds little, and concludes: "not notable".

That is not a neutral application of policy.

That is structural bias being mistaken for objectivity.

The same applies to books. A book published in English, French or German within a highly catalogued publishing ecosystem leaves a large institutional footprint. A book published in Swahili, an Indigenous language, or by a small publisher in the Global South may leave almost none.

Does that make the first book inherently more worthy of a Wikidata item?

Of course not.

It only means one society produced the kind of documentary trail that our editors know how to find.

And this is exactly why I find the direction of this reform so troubling. Instead of asking how Wikidata can better represent knowledge that existing systems have historically failed to document, you are proposing stronger mechanisms to reproduce those same omissions.

People are forgotten.

Scientists are forgotten.

Books disappear.

Local organisations disappear.

Entire communities can leave almost no digital trace.

Wikidata should be one of the tools that helps prevent that erasure, not a system that says: "You were not sufficiently documented by institutions we recognise, therefore you do not qualify to exist in our knowledge graph."

And again, what exactly is the philosophical basis for this distinction?

A murderer who receives intense international media coverage for a few weeks can easily become "notable" under conventional criteria.

A midwife who served an isolated community for fifty years may leave almost no written record.

Is the murderer more worthy of representation in human knowledge?

A fraudulent businessman covered extensively by newspapers may easily satisfy a notability threshold.

A scientist from an underfunded university in a developing country may not.

Does the first person deserve a QID more than the second?

This should make it obvious that notability is not importance.

Notoriety is not importance.

Media attention is not importance.

Digital visibility is not importance.

Institutional visibility is not importance.

These are proxies produced by particular social and economic systems, and they are distributed extremely unevenly across the world.

This is also why I fundamentally reject the premise that Wikidata must distinguish between "notable" and "non-notable" people or organisations in the first place.

If every reliably identifiable registered company were represented, no company could claim prestige merely from having a Wikidata item. A QID would simply identify an entity.

If broad classes of people were represented according to verifiability, privacy and ethical criteria rather than some vague judgement of importance, having a Wikidata item would likewise cease to imply that the community had declared someone "important".

It is selective inclusion that creates this distinction.

And then we use the distinction we created as justification for further exclusion.

For living people, we absolutely need privacy and data-protection rules.

For false information, we need verifiability rules.

For spam, we need anti-spam rules.

For promotional editing, we need conflict-of-interest rules.

For bad mass imports, we need mass-editing and data-quality rules.

For technical capacity, we need technical solutions and prioritisation.

But none of those problems requires us to appoint ourselves judges of which human beings, books, companies, organisations and cultures are important enough to exist.

I find it particularly concerning that a proposal with such profound implications for the representation of the Global South and marginalised knowledge is being developed largely through the epistemic assumptions and documentary practices of the Global North.

A brief "knowledge equity" exception does not fix that.

You cannot build a gate according to Global North standards and then claim equity because you added a small door for everyone else.

The burden should be reversed.

If an entity can be responsibly identified and described, those proposing its exclusion should explain the concrete harm caused by its presence.

"Not notable enough according to our unwritten standards" is not a sufficient answer for a global, multilingual and multicultural knowledge graph. Rodrigo Tetsuo Argenton (talk) 04:09, 23 July 2026 (UTC)reply

Thank you very much for putting my unease into writing. You are absolutely spot on regarding the list of problems to be resolved through specific policies, none of which require "us" (Wikidata editors) to make global value statements about the importance of specific items and the infrastructures of knowledge production any one society or community might operate under. Reinforcing the epistemic violence of the current technology stack and hegemonic knowledge ecologies is certainly not a path to embark on. OpenArabicPE (talk) 21:26, 14 August 2026 (UTC)reply

Some additional data

[edit]
Graph showing the number of Items created by all accounts and accounts with up to 20 and up to 50 edits in that month
Graph showing the number of Items deleted that were created by all accounts and accounts with up to 20 and up to 50 edits in that month

In light of the recent significant increase in promotional Item creation I have asked our data analyst to get some statistics. Here are graphs for

  1. how many Items per month are being created and how many of those were created by accounts with less than 20 edits and less than 50 edits within the month.
  2. how many Items per month are being deleted and how many of those were created by accounts with less than 20 and less than 50 edits within the month.

What I take away from these graphs and observing the increase in promotional Item creation: We are seeing an influx of Items by editors who have little interest in contributing to Wikidata beyond their own direct benefit and that are putting a strain on our admins. Lydia Pintscher (WMDE) (talk) 13:37, 31 July 2026 (UTC)reply

I agree with your overall point, but I think there's a confounding factor in the statistics. In my experience, there are many good faith editors who (probably participating in some outreach event) create a lot of items in a short period but don't bring any into a maintainable state, so the items end up getting deleted. They're trying to contribute, but they don't know how.
Also, thanks for the charts, but please consider using distinct colours or line styles, rather than just three greens. :)
Bovlb (talk) 15:33, 31 July 2026 (UTC)reply
Fair point on the colors. And yeah it is certainly not only promotional Item creation but this sudden steep increase I can't really explain otherwise. Lydia Pintscher (WMDE) (talk) 15:37, 31 July 2026 (UTC)reply
Thanks for the graphs Lydia Pintscher (WMDE) , very interesting. I would say the easiest way to solve this problem is to disable item creation feature from everyone under 500 edits. Or do you think the spammers would try to create 500 meaningful edits before they start spamming? After that the Wikidata editathons would be about how to edit existing old items rather than creating new. And while volunteers edit the existing items they will learn the culture and once they are given the right to create new items, they would not create so many items that are to be deleted. If this sounds too rough maybe the confirmed Wikipedia editors (extendedconfirmed, autoreview, editor, what else there is?) could be given the right to create items already before they hit the 500 edits milestone. Jjanhone (talk) 10:14, 3 August 2026 (UTC)reply
The issue with overly restricting new item creation is that improving existing items often requires new items. When i try to improve items, number one issue is unsourced statements. Best sources are books, and to add book as source via stated in i need to add an item for written work and edition. Then, if i want quality items for book, i probably also need an item for author and publisher, so everything turns into a campaign of creating new items. This is very different from wikipedias where small new contributions could be just edits to wikitext. So restrictions need to be careful in this area. MSDN.WhiteKnight (talk) 04:31, 5 August 2026 (UTC)reply

Items from archives

[edit]

Regarding items that would be considered notable based on archives, there's potential issue that some items described by archive entries are very small. For example, in Russian art archive the archive entry could describe exhibition announcement or couple of photographs. We probably want an item for artist described by archive entry, but not for every photo, document or artwork (including drawings). Is it possible to have consensus-based notability guidelines for concrete area, like Wikipedias have? MSDN.WhiteKnight (talk) 05:09, 5 August 2026 (UTC)reply

Proposal: request option for restricted item creation

[edit]

First, I should disclose that I offer consulting services to companies, organizations and individuals on Wikipedia and occasionally Wikidata. My comment necessarily reflects my own interests, but I believe my perspective has something to add.

The current proposed wording bans COI item creation entirely. Jjanhone has suggested conditional permission, and as much as I would personally enjoy this privilege, there seems to be little interest in this approach. So I'd like to propose something narrower, which keeps the prohibition intact, but offers a path forward.

The paragraph's closing sentence, as recently revised, now says if you have a COI, not to worry, a qualifying item "might" be created "eventually." The problem is obvious: no mechanism exists for the proposal and review of new items. But a fairly straightforward solution exists, and has been suggested previously:

  • Epìdosis first suggested a Notability/Questions page in Round 1, and it now appears in the draft itself for pre-creation questions.
  • Arlo Barnes suggested a board where non-conflicted editors could evaluate a topic on request.
  • Jasper Deng, while otherwise supporting the prohibition on COI item creation, wrote that "anyone with a COI should be able to request an item on the entity they want."

So there is some pre-existing support for this idea, it just hasn't been recognized as such. I would suggest replacing the final sentence of the paragraph with:

If you believe a subject covered by this restriction meets Wikidata's notability criteria, you may request its creation at Wikidata:Notability/Questions, including your disclosure per Wikidata:Disclosure of paid editing, the proposed statements, and the sources or identifiers that satisfy criterion 2.

This way, if an item is genuinely within the scope of Wikidata but the person proposing it is covered by this restriction, here is a process whereby it can be reviewed. From the community perspective, I think this is preferable to the current whack-a-mole, where an item created via COI turns up at requests for deletions and consumes admin time either way. The same venue could also be useful in situations like the one Emwille and ArthurPSmith describe, where an identifier has not yet become publicly visible. WWB Too (talk) 03:25, 13 August 2026 (UTC)reply

Knowledge equity guidelines

[edit]

I’ve been working on a guide that may be useful for considering knowledge equity when we need to distinguish the notability of an item. Perhaps one of the most common reasons for deletion requests is self-promotion, which is why this guide emphasizes its distinction from marginalized knowledge. For now, it’s an essay, and I imagine it will continue to evolve and may eventually become a guideline. Wondering if it might be helpful to link it in the “knowledge equity” section of the proposed policy. Of course, from there, the community should continue working on it both during and after the final decision on the notability policy. --Mariana Fossatti (WK?) (talk) 21:33, 17 August 2026 (UTC)reply

I think we should take this as a starting point to define "Knowledge equity" much more clearer than in the current proposal. Otherwise there will be inevitably people claiming that a SEO expert from Central Europe is being concerned by "marginalized knowledge" (@Kordishal, Mondo:). --Dorades (talk) 11:34, 5 September 2026 (UTC)reply
I don't know what you were reading, but I have not claimed in any way that the subject and item creator herself was concerned about marginalized knowledge. I simply commented to point out that due to her status as a polish woman in a male dominated field, that is likely to be of interest to some scholars now and in the future, she represent a small piece of marginalized knowledge. There are currently 25 male SEO specialists and 10 female SEO specialists on Wikidata.
Because of this I pointed out, that this represents a conflict between the policy on marginalized knowledge and self-promotion. As a result anyone discussing it should consider this before deciding to delete the item. However, I did not vote to keep, I simply commented, precisely because this could represent a problematic precedent for future Women who are SEO something and then claim this argument for themselves when trying to self-promote. Which probably would be completely a waste of time but still.
And personally I would like to read some scholarly output why SEO people are now targeting Wikidata for their snake oil products. And then it would be nice if women were similarly represented in such a study as men. And Wikidata could serve as a starting point to find people and resources. Kordishal (talk) 06:30, 6 September 2026 (UTC)reply
Sorry, I think I didn't make clear what I wanted to say here. In the context of Wikidata, "marginalized knowledge" does not refer to the commercial interests of a subject that, due to its geographical location, is not affected by marginalized knowledge at all. "Knowledge equity" is not a factor here at all. Please have a look at the linked document, especially Wikidata:Knowledge_Equity_Guidelines#Comparison:_self-promotion_vs_underrepresented_topics. --Dorades (talk) 10:18, 6 September 2026 (UTC)reply
I have read the document before, but didn't fully consider it the full definition of "marginalized knowledge" on Wikidata. I concede the point that by this somewhat strict definition her item does not fall into the category. As there exists a VIAF entry about her she is not in danger of being erased in the way that "marginalized knowledge" seeks to prevent.
That being said, she is still a member of a marginalized group and a reference on Wikidata could help with that, but that would have to be argued in a different way. This is likely more the area of wiki projects for women / gender gap and others that promote knowledge about marginalized groups and individuals and I am less sure if that should be used as a criteria to ignore the self-creation/self-promotion rule.
That being said the essay/guideline likely should include a clear distinction between the two cases. Kordishal (talk) 11:05, 8 September 2026 (UTC)reply

Round 3: Votes

[edit]
  •  Support This is definitely an improvement over what we have now, and I like the list of points to consider; in the end this will be a matter of community judgment on what belongs here. ArthurPSmith (talk) 14:28, 9 June 2026 (UTC)reply
  •  Support I approve the change for WD:N 2. At least they will make vanity edit more harder. :) --Fralambert (talk) 17:42, 9 June 2026 (UTC)reply
  •  Support Nicely balanced. - PKM (talk) 19:55, 9 June 2026 (UTC)reply
  •  Support because it improves on the present situation; I think in N2 having OR is fine because requiring "At least one external identifier or archival record" AND "At least one reference" would exclude from notability a lot of entities which are described by reliable printed sources but not (yet) in online databases having an external-id property in Wikidata. But at the same time, as I have said multiple times above (the last one in #Round 2: External identifiers: When do they support notability and when do they not), I remained concerned that just saying "At least one external identifier" without specifying which external IDs imply notability and which do not imply notability (e.g. ORCID iD (P496)) and how to determine which external IDs imply notability and which do not imply notability is a big issue, especially in perspective. So basically I think the new reformed policy, while restricting slightly the notability and making it slighly more clearer what is notable and what is not, still leaves a very big area of discretionality about which external-IDs imply notability and which do not imply it, and we will probably have to settle this issue soon in another long discussion like this. So I think this is a good step forward, but is still unsufficient to make the notability situation clear enough. --Epìdosis 22:08, 9 June 2026 (UTC)reply
    I will publish some data I have for this now in #Round_2:_External_identifiers:_When_do_they_support_notability_and_when_do_they_not in a bit to move the discussion about which ones should be supporting notability and which ones should not forward. Lydia Pintscher (WMDE) (talk) 10:19, 31 July 2026 (UTC)reply
  •  Support Ainali (talk) 10:07, 11 June 2026 (UTC)reply
  •  Oppose The proposal still does not include important caveats to make sure that it doesn't have an impact on underrepresented knowledge. Many people (including myself) have questioned the process to reach to this policy, that still doesn't include the people that might be most affected by it. Scann (talk) 15:18, 16 June 2026 (UTC)reply
    Do you have a suggestion for how to improve the situation and reach those people? So9q (talk) 17:12, 30 June 2026 (UTC)reply
  • I remain opposed to the inclusion of Wikimedia Commons in Criterion 1. This is because Wikimedia Commons has no notability criteria for categories, yet Commons categories are treated as equivalent to articles here on Wikidata. Consequently, the door remains wide open for users to effectively "game the system" and secure a data object via Wikimedia Commons.
Furthermore, in my view, Wiktionary is being incorporated incorrectly into Criterion 1. According to the proposal mentioned earlier, the "zero" namespace (i.e., dictionary entries) is supposed to be excluded; however, the current wording does not reflect this.
Otherwise, I am satisfied with the proposal. I would therefore vote with  Conditional support. --Gymnicus (talk) 19:45, 16 June 2026 (UTC)reply
Does this proposal use Commons as a positive example to emulate? This discussion is hard for me to follow, and that bit is especially bewildering. Jerimee (talk) 20:15, 5 October 2026 (UTC)reply
  • I like many parts of the proposal and I thank Lydia for the work she's doing, especially the section about marginalised knowledge, but I  Oppose it in its current form (I prefer not to use the {{Strong oppose}} template in order not to create problems 😉), due to the section "In order to avoid the appearance of a conflict of interest, creating an Item about yourself, your close relatives or your business is not permitted." Wikidata consists of over 50% scholarly metadata (Lubiana, Rasberry, Mietchen 2025), largely due to the WikiCite project (of which Lydia has been a main supporter), thereby supporting discoverability and reliability in the research ecosystem. By restricting researchers from creating content about themselves and their outcomes, the policy undermines scientific self-curation, which I believe is one of the platform's major advantages, especially in the age of AI. See also the recent online course for researchers (Burton, Areia, Bartov).—super nabla(🪰 msg) 00:02, 21 June 2026 (UTC)reply
  •  Conditional support, its better than the current policy, but still vague. 2.1.1 is my main concern. Currently, you see alot of fighting at RFD about what counts as "serious and publicly available references", but at least with this, people generally have the heuristics from enwiki that paid placements, press releases, and trivial mentions do not constitute serious and publicly available references. However, 2.1.1 will now add people arguing about what counts as an "independent authority" which seems to offer less clarity and have less clear heuristics than what counts as "serious and publicly available references." --IntensionalLogician (talk) 21:44, 22 June 2026 (UTC)reply
    +1 What is meant by independent? Independent from the contributor? From WMF? So9q (talk) 17:20, 30 June 2026 (UTC)reply
    Independent from the subject. Does anyone have suggestions for how to improve this? Lydia Pintscher (WMDE) (talk) 10:21, 31 July 2026 (UTC)reply
    Rather than have a blanket policy for this, it would probably be better to discuss if the issuing authority is independent on a per external identifier basis. We could broaden Wikidata property for an identifier that suggests notability (Q62589316) and use it on any properties that have consensus as independent authorities. When proposing a new property, there could be a section where the proposer indicates if it the issuer is intended to be independent. Existing external IDs can be handled as they come up, but there needs to be a discussion around government issued ones like SIRET number (P3215) which anyone can get with a bit of work. -- IntensionalLogician (talk) 18:53, 31 July 2026 (UTC)reply
  •  Oppose I feel I need to oppose this, and appreciate and agree with the comments above by Super nabla and Scann. This decision is informed by my own past experiences, which I'd like to share. I came to Wikidata after leaving Wikipedia because I felt strongly that their notability policy was weaponized by certain editors to delete articles about historically marginalized groups and people, and ultimately drive editors away who wished to create articles about those people. I've also created items for authors after creating Name Authority Records (NAR) in Library of Congress' database for them, and had at least two I can recall deleted. One because the NAR had not yet reached the public interface and the link did not resolve, and one because I waited until the link would resolve before adding it. At times this can well exceed 24 hours (we have a government shutdown problem sometimes). I've also had an author item deleted that I created because the person was the author of an article. None of these were the end of the world, and the editors restored them after I contacted them and explained the situation. But I have never been contacted about an item I created before it was deleted, and even at my comparatively very low output these were frustrating. I worry that by adding more criteria and more potential hoops for people to jump through, we will discourage item creation and new editors, especially by and about the members of marginalized communities we want to include. I sympathize with editors that struggle with unclear criteria in deciding if an item is notable or not, but I worry that implementing these changes will cost us more than we will gain. Apologies for typing a novella at you, and thank you everyone for your work and discussion. --Emwille (talk) 20:06, 25 June 2026 (UTC)reply
    There is a good point raised here - items may be added to Wikidata at the same time they are being added in external reliable sources like this, but before those additions are visible to the outside world. Is there any way we can have a standard process for indicating that this item refers to an entity that will soon be available in a standard reference source, but is not there yet? I can imagine many situations where this will come up, not just for marginalized groups. ArthurPSmith (talk) 20:44, 27 June 2026 (UTC)reply
    We could have a new "waiting" namespace or status on items? So9q (talk) 17:15, 30 June 2026 (UTC)reply
    Which situations would this be? Are there any where we can not wait a bit? Lydia Pintscher (WMDE) (talk) 10:22, 31 July 2026 (UTC)reply
    Do you mean "where we can not wait a bit [until an external identifier is created to create the Wikidata item]?" I don't want to put words in your mouth, just trying to clarify what I'm answering. If it was a different question feel free to disregard.
    My current workflow for creating an NAR and Wikidata item:
    1. Create the WD item
    2. Create the Name Authority Record [NAR] and add the Wikidata item in an 024
    3. Submit the NAR to the authority file and forget about the whole thing
    4. LCCNbot adds the LCCN from the NAR to the WD item for me (thanks!)
    A new workflow where I waited a bit would probably look like:
    1. Create an NAR
    2. Wait a few days to see if it's made it out of distribution to the public side.
    3. There's no notification so I have to manually check every time.
    4. If there's a government shutdown who knows how long it will take to come out of distribution.
    5. Once it's out of distribution, create a Wikidata item and include the LCCN in the WD item.
    6. Go back in to the LCCN and edit it to include the Wikidata item in an 024 field (if no one else has edited it and locked it for distribution again) and resubmit to the authority file.
    That might not sound like a whole lot more, but it severs one process I can do simultaneously into two separate processes where I have to keep checking to see if I can do process two, and in reality I'll just stop creating Wikidata items for corresponding NARs. I'll still do other stuff, but I do this as a volunteer. No one in my reporting line will care or probably even notice if I stop, and making it more difficult for me is going to be a good enough reason to take it off my plate.
    And I think that needs to be a factor overall. You're building a taller fence to keep people out because they're submitting items that don't meet the notability criteria, but how many people are you going to accidentally drive away because they get an item deleted on Monday because it doesn't have an external identifier publicly available until Tuesday? Emwille (talk) 20:33, 17 August 2026 (UTC)reply
    Hi, I think my experience is similar in some aspects so I share it! I have editing access to the authority file of OPAC SBN (Q105086090) (i.e. SBN author ID (P396)), where I can edit existing authority records or create new ones; the backend is copied into the public interface once a week (or two weeks, in a few cases of national holidays, or unforeseable technical issues); so while I am certain my edits are approved, they might be visible after a few days, usually no more than 7. It has happened to me, in a few occasion, to follow a worflow very similar to the one of @Emwille::
    1) create a new SBN ID
    2) create WD and put in it the SBN ID [the item always has also other sources, but sometimes its notability rests mainly on the presence of P396]
    3) after some days, the ID SBN becomes visible
    So in my case, the issue would mainly be the following: if, when I create a new SBN ID, I have to wait until it becomes visible to create the corresponding Wikidata item (because, if the SBN ID is not yet visible, the item might not be notable), the procedure becomes more complex because I have to annotate somewhere that I should create a Wikidata item for a certain SBN ID and come back after some days to create it; doing everything in the same minutes is more efficient, the other option, as Emwille very well said, "severs one process I can do simultaneously into two separate processes".
    That said, I have a few more reflections:
    A) specifically on the NAR, because I don't know exactly how this aspect works: when you submit the NAR to the authority file do you get assigned an ID which then becomes visible after approval or the ID is assigned only when the approval happens? In the case of SBN, I see the ID in the moment in which I create the ID, so I can instantly use it in Wikidata, although if clicked the ID would lead to a blank page until the weekly update of the public interface happens; if you have the ID instantly, I would suggest you to add it in Wikidata just after the creation (although still inactive) to demonstrate its existence, because
    B) in my experience, the items I have created whose notability relied on a not-yet-visible SBN ID have never been questioned (probably because no one opened them in the few days before SBN was visible), so I expect the same would happen in similar cases for items based on NAR; and, in general, I think that if we know that a not-yet-valid ID has been added by a user able to create IDs in the same database, and the user says the ID will soon be visibly valid, accepting the item is the most reasonable solution for us;
    C) if, conversely, NAR assigns the ID only after validation, I would suggest one of these two possibilities: edit "they meet at least one of the above criteria within 24 hours" enlarging to some days (one week?) the timespan to check if the item reaches notability, which could avoid issues for these items but possibly creating more serious issues in other fields; or just adding Library of Congress Authorities ID (P244) with novalue and a qualifier has characteristic (P1552) with value "record created but not yet visible" or something similar, and then instructing @LccnBot: to delete this when it adds the definitive ID. Epìdosis 21:32, 17 August 2026 (UTC)reply
    I can confirm that the identifier for a Library of Congress Name Authority record is available immediately after the record is submitted. However, it's very easy to forget to add the identifier into Wikidata after submitting the record, which is why we created Lccnbot. Mcampany (talk) 23:06, 17 August 2026 (UTC)reply
    Yep! I would also add that after I had a Wikidata item deleted because the LCCN was not *publicly* available yet (it was available to me and I put it in the Wikidata item) and when someone clicked on the link in the Wikidata item it did not go to a valid LCCN I was happy to let LCCNBot do the work because a. I'm happy to let bots handle it when they can and b. I thought a link added by a bot might be seen as more reputable and less likely to get an item deleted. Has LCCNbot had any issues you know of with its edits and associated items being deleted by editors because the LCCN isn't publicly available yet?
    I should also add the editor who deleted the item willingly restored it quickly after I explained the situation, which I do appreciate. It was still frustrating though. Emwille (talk) 00:18, 18 August 2026 (UTC)reply
    Hmm... I don't think I've run into that issue yet, and I'm not sure what would happen if the bot tried to add an LCNAF identifier to an item that no longer exists. @Thisismattmiller would know, though. He does all of the heavy lifting coding-wise. Mcampany (talk) 00:36, 18 August 2026 (UTC)reply
    The API calls will fail if it tries to add a P244 to a deleted item. So nothing will happen. I looked in the last year of edits by the bot there were ~30 times it added a p244 and someone removed it, all for normal maintenance reason looks like. Thisismattmiller (talk) 20:59, 18 August 2026 (UTC)reply
  •  Support A marked improvement over the current policy. William Graham (talk) 14:54, 27 June 2026 (UTC)reply
  •  Oppose The phrasing under other options is a bit misleading. I would prefer something like this instead:
    If your data does not fit Wikidata's notability criteria, you may wish to consider creating or contributing to another Wikibase instance. Wikibase Cloud and Wikibase Suite provide software for hosting independent knowledge bases that can support data outside Wikidata's scope. While these projects are part of the broader Wikibase ecosystem, they are separate from Wikidata, and integration between independent Wikibase instances and Wikidata is currently limited.
    or
    If your data does not fit Wikidata's notability criteria, another Wikibase instance may be an appropriate place for it. However, independent Wikibase instances are separate projects from Wikidata, and users should not expect integration, discoverability, or interoperability with Wikidata.
    --So9q (talk) 17:07, 30 June 2026 (UTC)reply
  •  Oppose see discussion. Jjanhone (talk) 18:33, 7 July 2026 (UTC)reply
  •  Conditional support I oppose if wikisource subpages are not considered, because wikisource transclusion has a maximum limit. So pages containing recipes like those on s:id:Mustikarasa/Tortilla cannot created wikidata items. Empat Tilda (talk) 17:44, 12 July 2026 (UTC)reply
  •  Support That's a tough piece of work with a nice outcome. Not perfect, it never is, but it's good policy. --Denny (talk) 12:46, 15 July 2026 (UTC)reply
  •  Support -- Ooligan (talk) 05:28, 18 July 2026 (UTC)reply
  •  Support We need to reform our policy due to the fact that today's Internet is no longer the Internet we knew 14 years ago, when this project is started. I think that, however imperfect, this proposal is strongly addressing the problem of spam and self-promotion, which is a growing problem on the platform. I would be also in favour of So9q's phrasing for alternatives, but I can do without. --Sannita - not just another it.wiki sysop 19:16, 18 July 2026 (UTC)reply

 Strong oppose

I strongly oppose this reform in its current form, and not merely because of one particular clause.

The proposal takes real problems, such as spam, self-promotion, poor-quality mass imports and maintenance burden, and attempts to solve them by restricting what knowledge is allowed to exist in Wikidata. These are different problems and should not be conflated.

Wikidata is not Wikipedia. It is a general-purpose knowledge graph. An entity does not need to be "important" in an encyclopaedic sense to be useful as structured data. If an entity can be clearly identified, reliably distinguished from other entities and supported by verifiable information, the burden should be on demonstrating why its presence is harmful, not on forcing it through an increasingly subjective test of "notability".

This proposal risks formalising precisely the kind of systemic bias that Wikidata should be reducing. What counts as "significant coverage", a "serious" source, an "independent authority", or even an entity sufficiently important to deserve representation is not culturally neutral. These concepts strongly favour subjects documented according to the institutional, publishing and digital practices of wealthier countries.

A company in Europe, a book published in English, or a cultural institution in North America is far more likely to have the kinds of sources, identifiers and digital footprints that editors recognise immediately. A local organisation in Pirapora do Bom Jesus, a publication in Swahili, an Indigenous cultural entity, or an institution in a poorly digitised region may be perfectly real, identifiable and relevant while leaving a completely different documentary footprint.

The proposed "Knowledge equity" paragraph does not solve this. A single sentence saying that marginalised knowledge will be "taken into consideration" does not compensate for adopting criteria that structurally create the problem in the first place.

I also reject the idea that technical scaling, spam or administrator workload should be solved through epistemic exclusion. We should address mass creation through mass-editing policies, poor data through sourcing and modelling requirements, and promotional behaviour through conduct and conflict-of-interest rules. None of those requires us to import a Wikipedia-like conception of notability into a general-purpose knowledge graph.

Before adopting a substantially more restrictive policy, we should have evidence that the current inclusion criteria themselves are the cause of the stated problems, an assessment of which existing classes of items would be excluded, and an impact analysis across languages, regions and knowledge systems.

We do not have that.

The proposal should therefore not be adopted in its current form. Rodrigo Tetsuo Argenton (talk) 03:56, 23 July 2026 (UTC)reply

This would imply that every single living person and every single company on the Earth is notable and can have a Wikidata item, since there are sources confirming their existence. I do not think there is community consensus that this would be an acceptable development. Ymblanter (talk) 11:00, 24 July 2026 (UTC)reply
Living people 'only' constitute eight (or nine in a decade or so) billion records, which is a lot but not in theory impossible to record. But of course, the majority of those people not only are not already recorded in a public manner but wouldn't want to be. And there would be little benefit to recording someone whose documentary influence doesn't extend beyond a limited locale consisting of their immediate community, and potentially significant harm, especially if they are not consulted. On the other hand, someone who does want to be in the database may either have benign intentions, like adding context to an existing item, or may simply wish to self-promote. Whether this itself should be considered harmful depends on what we consider Wikidatans to be WD:HERE to do. Arlo Barnes (talk) 18:51, 31 July 2026 (UTC)reply
I think Rodrigo is scared of having sources verified by a non-local and having to cite online sources. Neither is a requirement of the proposal. Wikidata has properties for citing a book. If you want to make a proposal to dissalow people that do not speak the language that a source depicts to verify it, then I will gladly support that. Snævar (talk) 16:46, 15 September 2026 (UTC)reply

 Oppose Hi, I am pretty new here, but I would still like to put my opinion on this. When I first started to edit Wikidata, I had felt it odd that there is a notability policy in the first place. Now since it was very permissive you could ultimately argue for pretty much anything to be added, but clearly that has led to many conflicts between those that maintain Wikidata and those that wish to add new items to Wikidata. I understand why this policy update is being drafted, and I do believe it would be useful in maintaining Wikidata. However, it seems to me that there is a fundamental conflict between wanting notability to ensure maintainability and correctness and wanting to help spread and preserve marginalized knowledge.

To me it appears the primary purpose of this change in policy is to make it easier for administrators and editors to decide if they should delete an item or not. Making it harder to add marginalized items is a side effect of that purpose.

One part of this is that the essay in maintainability possits that shared stewardship requires any person to be able to verify what an item is. I don't disagree with the sentiment, but it is impossible for all knowledge to be verifiably by all people. An editor in Londen will never be able to check if a person was actually mentioned in some physical archive in a small town somewhere in Senegal.

So either the policy should be changed to ensure that only items can be added that can be digitally verified from "authorized" sources and acknowledging that this excludes knowledge about marginalized topics and entities or it should be kept as permissive as it currently is to ultimately allow any kind of knowledge to be added to Wikidata. Only the latter would allow Wikidata to fill a role of preserving and dissiminating marginalized knowledge.

A second issue to me is, that the question on whether or not something should be on Wikidata should not be "Is this entity notable?", instead it should be "Is this useful to a project or purpose that Wikidata is supposed to support?".

The first rule of the current policy answers the second question, because Wikidata is used as interwiki link. As a result any items created from these links are useful to Wikidata, because they help fulfull its primary (or at least one important) purpose. That many of these items fullfil the criteria of notability from their host wiki is only circumstancial.

The third rule also answers this second question, because they are useful to describe other items on Wikidata. This is desirable since this is supposed to be a knowledge graph and having more connections is generally a good thing.

The second rule on the other hand just states, that it has to be something that exists and that someone has written about somewhere. While the policy change attempts to specifiy what this means, I don't think it is a fundamentally useful question for an open knoweldge graph. Any inclusion of data in Wikidata has a goal, that may be tracking knowledge from other knowledge graphs (like authority files, geonames, gene data bases), dissimenate knowledge about research, self-promotion etc. It would seem more useful to me to track the goals of entities and connect them to a process outside of Wikidata and then decide deletions policies based on that. This would for example ensure that the open knowledge graph could absolutely be used to support open science and scholars in dissemenating their research. Of course I don't know how this could be exactly implemented, and it would definitely not be possible to decide something like that here.

  •  Support After I've been reading some delete stuff, others opinions, familiarizing myself with the knowledge equity guidelines and more I think I can see that the risk of deleting marginalized knowledge is fairly low. Deletion of knowledge about marginalized subjects is still at risk, but this is difficult to host on Wikidata anyway since we can't be a place to host original research. So I am changing my vote to support as well. A minor adjustement I would make is to clarify the text in N#2.2 as it first speaks of just one source, but then from multiple sources. I think it should me made clear that one source is sufficient, but multiple are preferred. A potential name change will need to be discussed separately.

Kordishal (talk) 10:52, 9 August 2026 (UTC)reply

Thanks for your comment.
I don't think anyone is arguing that maintainability depends on every editor being able to access every source, or that only online sources are acceptable. In particular, WD:MAINTAIN offers the much weaker position that it should be possible for some editors other than the item creator to maintain the item. In my experience of reviewing item creations, the real frontier is at the other end of the spectrum: whether any sourcing or external identification is provided at all.
I agree that we should avoid making it unnecessarily difficult to preserve marginalized knowledge, but Wikidata is not designed to be a primary source. If Wikidata is the only place in the world where an entity is described or a claim is documented, then other editors have no independent basis on which to maintain that information. We need to accommodate knowledge documented in difficult-to-access sources without abandoning the requirement that there be some basis for shared stewardship beyond the knowledge of the item creator. Bovlb (talk) 14:07, 9 August 2026 (UTC)reply
To be honest I probably read your Maintainability essay stricter than you meant it. It was a pretty specific.
Do you think it would be possible to kind of create an overview over how many items would be deleted if this change in policy was applied strictly? At the very least that would be items with no external identifiers, no qualifiers on any statements with external urls of any kind, and no statements with links to other items that have some kind of authority by themselves. It wouldn't be trivial and not fully representative, but potentially such a dataset could maybe be used to determine if there is a set of items that are not at all covered by a strict reading of this new policy. And just to be clear I don't think there is any intention of doing automated deletes in such a manner in response to this policy reform.
And to maybe be a bit more concrete:
On policy 2.2, am I understanding that correctly that the first sentence means an item must have at least one independent reference, and the rest of the point is (beyond an explanation of what a valid reference is in this context) that there should be more than one reference? If yes, then maybe the policy could be a bit more explicit about this.
Then to be a bit more constructive about the concerns around marginalized knowledge. I think that the current disclaimer on knowledge equity puts all the responsibility of getting it right on the deletion process of an item. I am not currently familiar with that process in detail, but do you think there are resources and options in that process that ensure that people from marginalized communities can make their case? And that admins/editors who arbitrate deletions have the resources to make informed decisions? I think it would be useful to expand the section on Knowledge equity with concrete measures or links to concrete policies that are intended to prevent marginalized knowledge from being deleted for the wrong reasons. Instead of the current verbal only affirmation.
That being said, I absolutely agree with you that Wikidata cannot be a primary knowledge-base for any kind of knowledge. So some kind of external reference is always needed. We just need to make sure that these references are respected even if they are not in line with Western / Global North standards.
And lastly I really do think Notability is kind of a bad name for this policy. It immediately has a judgemental (wertend, not sure if that is a good translation) tone to it in my opinion which gives the impression that any knowledge that is not on Wikidata is not worthy of its consideration. A more neutral name like Inclusion Criteria or similar would much better serve the community, as it makes it clear that some knowledge is excluded for valid reasons, and not just because "we" (some nebulous wiki-community) decided its not worthy. Though I don't know if a name change is in the scope of this discussion or if that would need to be a separate process. Kordishal (talk) 07:05, 11 August 2026 (UTC)reply
Looking at items created over a one-week period, ending one week ago, using a mechanical application of the notability policy, 1.5% still have no detected indication of notability. A further 2.6% have been deleted (2.1%) or redirected (0.5%), and might also have lacked such indications. This gives an overall range of 1.5–4.1%.
I agree that “inclusion criteria” may be a clearer name for this policy for many newcomers. A common stumbling block I see is the distinction between the real-world significance of an entity and the evidence provided by its representation in Wikidata. Our notability criteria focus on the latter. Bovlb (talk) 17:11, 11 August 2026 (UTC)reply
Oh that is a very nifty tool! Quite useful to get a quick overview. I'll definitely let you know should I find any issues, but my random jumping around was quite successful so far.
One note https://wd-notability.toolforge.org/?qid=Q100706126, I don't think Google Knowledge Graph ID should be used as a inclusion criteria? At least not by itself, or does it not matter because its always with a wikipedia entry? Or does it currently just use all external ids to indicate STRONG 2.1?
Kordishal (talk) 09:38, 12 August 2026 (UTC)reply
The tool does does consider Google Knowledge Graph ID (P2671) to be a STRONG identifier; it's an identifier and it's not social media. I think the list could use a bit of tweaking. I hope to externalize the configuration of the tool soon. Under the "AND" interpretation of N2, this only satisfies N2a, so only raises the item to "PARTIAL-STRONG". This item also has a sitelink, making it overall STRONG. I suggest that more detailed discussion of the tool continues at its talk page. Bovlb (talk) 16:52, 13 August 2026 (UTC)reply
  •  Support I support this proposal --Uschoen (talk) 19:52, 11 August 2026 (UTC)reply
  •  Oppose due to the way #2 is written. This makes #2 contradictory and useless. I support stronger restrictions on COI item making, imo should be totally disallowed. PARAKANYAA (talk) 02:48, 13 August 2026 (UTC)reply
    •  Support On second glance, I believe I misinterpreted this. Provided 2 is an OR, not an AND, on requiring these things (e.g. one serious, independent, reliable source OR serious notability proving identifiers like an LC authority, not both), I would support this. I would fiercely oppose it being AND; plenty of very notable people who are described in full length books do not have identifiers. PARAKANYAA (talk) 02:52, 13 August 2026 (UTC)reply
  •  Support, as I stated in the Project chat regarding the Deletion policy and regarding the quality of media files; my essay regarding "Do not disrupt Wikidata to illustrate a point"; Wikidata:Disclosure of paid editing which is a policy; Wikidata:Self-promotion which is currently an essay; and Wikidata:Autobiography which doesn't have a banner, so I assume than it is an essay. --Amitie 10g (talk) 20:33, 8 September 2026 (UTC)reply
  •  Support There has been a need for higher requirements for a decade now. As a citizen of one of the smallest countries in existance I reject arguments of this limiting information to whatever is widely available, as that argument really does not have any merits other than being a personal opinion that has no basis in facts.--Snævar (talk) 16:30, 12 September 2026 (UTC)reply
  •  Oppose I disagree that Wikidata's purpose has changed in a way that warrants a overhaul of the notability standards. This proposal seems heavily influenced by the norms of a few highly active projects that have dedicated teams enforcing strict local rules. Restricting item creation based on how proficiently a user adds sources is like trying to manage population growth by restricting who can have children, it kind of targets the wrong issue. So instead of blanket restrictions or "efficiency for efficiency's sake", creating a detailed collection of "what we dealt with in deletion requests" would be more helpful in decision-making. For example, a regional government agency in a highly populous country might serve millions of people, yet it will risk deletion. Or someone can simply unlink on the local wiki, leading it to the deletion on Wikidata. Or the fact that communities do not have the manpower to police spam or verify content at the same pace. The community of Wikidata is bringing together editors and easing the work of local wikis, and it seems to have done a great job and it can still do a great job on the "old" 2013 notability rules. Profesorul Blazat (talk) 22:35, 7 October 2026 (UTC)reply
 Support. This is a major improvement to the current policy. I see the point on knowledge equity, but I don't think this change sets the bar too high to seriosly hamper it. The 2nd criterion could be formulated slightly more clearly by replacing "The minimum criteria:" with "if it meets at least one of the two subcriteria below:". Samoasambia ✎ 12:09, 12 October 2026 (UTC)reply
Retrieved from "https://www.wikidata.org/w/index.php?title=Wikidata:Requests_for_comment/Notability_policy_reform&oldid=2557274389"