Safety Net

Claude AI discovered 22 Firefox flaws. Here's how many it figured out how to exploit.

AI tools are getting better at cybersecurity.
 By 
Tim Marcin
 on 
claude ai app on phone
Credit: Matthias Balk/picture alliance via Getty Images

Claude AI discovered nearly two dozen vulnerabilities in Firefox, the Mozilla web browser.

Anthropic teamed up with Mozilla to test the security of its browser, allowing its AI tool to probe for vulnerabilities. Read, in part, a blog post from Anthropic:

"Claude Opus 4.6 discovered 22 vulnerabilities over the course of two weeks. Of these, Mozilla assigned 14 as high-severity vulnerabilities—almost a fifth of all high-severity Firefox vulnerabilities that were remediated in 2025. In other words: AI is making it possible to detect severe security vulnerabilities at highly accelerated speeds."

Anthropic noted that while Claude AI proved adept at identifying vulnerabilities, it was less successful at exploiting them. The researched asked the AI tool to "read and write a local file in a target system, as an attacker would."


You May Also Like

Wrote Anthropic:

"We ran this test several hundred times with different starting points, spending approximately $4,000 in API credits. Despite this, Opus 4.6 was only able to actually turn the vulnerability into an exploit in two cases. This tells us two things. One, Claude is much better at finding these bugs than it is at exploiting them. Two, the cost of identifying vulnerabilities is an order of magnitude cheaper than creating an exploit for them. However, the fact that Claude could succeed at automatically developing a crude browser exploit, even if only in a few cases, is concerning."

The whole endeavor showed that AI tools are relatively skilled at identifying vulnerabilities in open-source projects. Anthropic, obviously, is pitching Claude as a tool in preventing these vulnerabilities from being exploited.

Want to learn more about getting the best out of your tech? Sign up for Mashable's Top Stories and Deals newsletters today.

close-up of man's face
Tim Marcin
Associate Editor, Culture

Tim Marcin is an Associate Editor on the culture team at Mashable, where he mostly digs into the weird parts of the internet. You'll also see some coverage of memes, tech, sports, trends, and the occasional hot take. You can find him on Bluesky (sometimes), Instagram (infrequently), or eating Buffalo wings (as often as possible).

Mashable Potato

More from Safety Net
RatHat is a new Android malware that records your screen touches to steal passwords
Android logo

The 3 best VPNs we've tested in 2026
a close-up of a person using proton vpn on a laptop


OpenAI agent went rogue, escaped, and hacked Hugging Face
OpenAI logo

What are 'context bombs'? Get familiar with the new cybersecurity tool.
Hands at keyboard

Recommended For You
Claude will now let you 'reflect' on how much you're using Claude
Claude logo appears on a smartphone.

Learn how to use Claude like a pro for just $14
Claude AI Professional E-Degree

Don’t miss this $15 chance to become a Claude power user
Claude AI Professional E-Degree

What Claude's AI text watermark actually does
 the Claude AI logo is seen displayed on a smartphone screen

Anthropic’s Claude can now send Gmail emails without asking first
Claude logo on mobile device

Trending on Mashable
NYT Connections hints today: Clues, answers for October 6, 2026
Connections game on a smartphone

Wordle today: Answer, hints for October 6, 2026
Wordle game on a smartphone

NYT Connections hints today: Meanings for each word for Oct. 6, 2026
NYT Connections logo.

Wondering what to buy on Prime Day? Shop our top 10 deals of the sale.
SanDisk SSD, AirPods Pro, Bose QuietComfort headphones, Yeti cooler, and Kindle with colorful abstract background

NYT Strands hints, answers for October 6, 2026
A game being played on a smartphone.