What Are Contact Centers Actually Doing With AI?

Contact center customers are starting small with AI. They pick a narrow, high-volume job, such as answering missed calls, resetting passwords or analyzing call recordings, prove the return and then buy more. That pattern runs through the latest earnings reports from multiple CX vendors. Adoption is broad by customer count and shallow by share of conversations. The complex, multistep work vendors talk about most is the work customers have done the least.

The vendors are seeing AI show up in revenue. Among those that disclose a comparable figure, AI accounts for roughly 12% to 15% of revenue, depending on what it's measured against. No vendor breaks out how much AI contributes to profit.

For CX leaders, the more useful details came from executives describing what customers are actually doing with AI: where they start, where they stall and where they report results.

So we'll go there.

How CX Vendors Measure AI Revenue

Five9, NiCE and RingCentral each reported an AI figure at about 15%, but each measures something different. Five9 counts revenue from AI products. NiCE counts AI annual recurring revenue (ARR), including the Cognigy business it acquired. RingCentral counts the total ARR of customers who bought at least one paid AI product, which is the broadest definition of the three.

How CX Vendors Report AI Adoption

Editor's note: Figures are company-reported and come from each vendor's most recent earnings report or results announcement as of Oct. 28, 2026. Fiscal calendars differ, so the months each report covers are listed. Genesys is privately held, and its figures are unaudited.

Vendor	Earnings Report	AI Adoption Figure	Source
RingCentral	Q3 2026 (July–Sept. 2026), reported Oct. 22, 2026	Customers using at least one paid AI product account for about 15% of total ARR, more than $400 million, doubling year over year. This counts those customers' full spend, not AI revenue alone.	Earnings release
8x8	Q1 FY2027 (April–June 2026), reported Aug. 4, 2026	The number of customers adopting 8x8's AI solutions rose 121% year over year. Usage-based revenue, which includes communication APIs, AI solutions, digital channels and telecom, grew 63%.	Q1 FY27 metrics release
Five9	Q2 2026 (April–June 2026), reported Aug. 6, 2026	AI revenue was 15% of subscription revenue, up from 9% a year earlier. AI revenue reached $39 million in the quarter, up 78%.	CMSWire
Genesys	Q2 FY2027 (May–July 2026), announced Sept. 2, 2026	Genesys Cloud AI ARR topped $400 million, growing more than twice as fast as Genesys Cloud ARR, which neared $2.9 billion. By CMSWire's math, AI is roughly 14% of Genesys Cloud ARR.	Company announcement
NiCE	Q2 2026 (April–June 2026), reported Aug. 5, 2026	AI ARR reached $362 million, 15% of cloud revenue, up 52%, down from 66% growth the prior quarter. The total includes Cognigy.	CMSWire
Sprinklr	Q2 FY2027 (May–July 2026), reported Sept. 2, 2026	ARR from AI-native SKUs grew 40% year over year, with more than 200 AI engagements underway. Sprinklr did not disclose a dollar figure.	Earnings call transcript
Twilio	Q2 2026 (April–June 2026), reported Aug. 6, 2026	Self-serve voice, the main entry point for AI-native customers, grew more than 50% year over year.	CMSWire
Zoom (Zoom CX)	Q2 FY2027 (May–July 2026), reported Aug. 25, 2026	Zoom Virtual Agent customer count rose 256% year over year. Zoom CX ARR grew at a high-double-digit rate, and paid AI was part of nine of the top 10 Zoom CX deals.	Earnings release

Where Customers Are Adopting AI

Voice Self-Service at the Front Door

Now the good stuff if you're a CX practitioner trying to make sense of AI deployments.

Voice is the most common entry point. RingCentral's AI Receptionist (AIR) has more than 22,000 customers, up nearly 4x year over year. CFO Vaibhav Agarwal said on the Oct. 22 earnings call that customers use it to address missed calls and capture more leads, and that it sells mostly to small businesses today while moving into mid-market and enterprise accounts.

At Twilio, self-serve voice grew more than 50% in the second quarter. CEO Khozema Shipchandler said voice carries most AI interactions for now.

"Most of the interactions that we're seeing today are on the voice channel. We would expect that a lot of those move over to other channels over time, too, which also generates pretty good durability for the business," Shipchandler said.

8x8 reported voice AI interactions grew more than 106% year over year.

Simple, Repetitive Requests First

Five9 CEO Amit Mathradas said customers deploying its Voice AI Agents start with simple use cases like password resets, then expand into more complex, regulated scenarios, CMSWire reported in August. At NiCE, CEO Scott Russell cited GX Bank in Malaysia, where AI resolves 70% of chat interactions on its own.

Zoom sells Zoom Virtual Agent both with Zoom Contact Center and as a standalone product. Its customer count rose 256% year over year.

Conversation Analytics and Quality Management

Analyzing conversations is the other common starting point. RingCentral's AI Conversation Expert (ACE) is approaching 7,000 customers, up 60%. On the earnings call, President and COO Kira Makagon described customers using it to identify call drivers and to grade and coach agents.

Sprinklr said it's seeing outsized growth in its agentic and contact center intelligence products. Genesys reported in March that the number of organizations using its Supervisor AI capabilities more than doubled during fiscal 2026.

Agent Assist and Platform Attach

Genesys said more than 70% of Genesys Cloud customers were using Genesys Cloud AI as of Jan. 31, 2026, and AI-powered conversations on the platform grew more than 120% during the fiscal year, according to its fiscal 2026 results. AI made up 20% of new Genesys Cloud business annual contract value.

Zoom said paid AI was part of nine of its top 10 Zoom CX deals in the quarter. CEO Eric Yuan said customers are increasingly combining Zoom's virtual agent with AI-assisted human agents.

Small Businesses and Lightweight Contact Centers

Smaller customers are adopting quickly. RingCentral's Customer Engagement Bundle (CEB), a lightweight contact center for businesses without a formal one, passed 14,000 paying customers, growing nearly 50% quarter over quarter. Nearly two-thirds of CEB customers have bought at least one AI product. 8x8 said the customer base for 8x8 Engage, its product for frontline and non-desk workers, grew more than 247%.

RingCentral also named healthcare and financial services as its strongest verticals for AI.

What's Lagging When it Comes to AI in CX?

AI Isn't Handling Most Calls Yet

Customer counts are rising faster than AI's share of actual conversations. RingCentral CEO Vlad Shmunis said so directly on the earnings call.

"I can tell you that our traffic growth has been strong but linear. Our AI growth has been explosive, obviously. But overall penetration of AI and AI-answered calls on our network is still very small percentage-wise," Shmunis said.

Twilio's Shipchandler called voice AI "very early innings," noting most people have had fewer than five or six voice AI interactions in the past year. Five9 said its concurrent contact center seats are still growing, with no seat compression.

"We are still in the early stages of a much broader AI adoption cycle across our customer base," Russell said in NiCE's earnings release.

Complex, Multistep Workflows

Autonomous agents that handle multistep work are the next phase for most vendors. RingCentral's AIR Pro, Five9's regulated use cases and Twilio's human-to-agent interactions are all described as expanding from a smaller base. Shipchandler said fully autonomous agent-to-agent interactions are still "a while" off, particularly in financial services and healthcare, where a human in the loop remains necessary.

Enterprise Deals and Partner Channels

Five9's largest deal of the quarter, a roughly $100 million contract with a Fortune 100 financial services company, covers the contact center platform only. AI components are still to be determined. RingCentral's Shmunis said AI adoption through carrier and telecom partners lags direct sales.

"When we say that it's still pretty early, very early with AI, in general, for this type of an application, it is yet earlier through the channel and in particular with GSPs. They just move a little slower, but once they move, they move with very strong force," Shmunis said.

At NiCE, AI ARR growth slowed to 52% from 66% the prior quarter. One analyst estimated that without Cognigy, NiCE's AI growth was below 30%, CMSWire reported. NiCE said the acquisition can't be separated out because it discontinued comparable products when it bought Cognigy.

Deployment Capacity and Costs

Getting AI live takes people, and running it takes compute. Five9's adjusted gross margin fell to 61% from 63% a year earlier after it temporarily added professional services staff to deploy AI customers faster than planned. Sprinklr's services business ran a negative 22% non-GAAP gross margin, and the company said higher data and hosting costs came with expanded AI usage.

"While we delivered solid results, we recognize that our professional service and support organization requires greater focus," Sprinklr CEO Rory Read said on the earnings call.

Zoom also said AI usage spiked with new products, raising expenses in the quarter.

Where Customers Report Results 

Expanding After the First Product

RingCentral said a national fitness operator started with ACE to identify call drivers, then added AIR for recurring inquiries and expanded its investment 4x. Agarwal said customers using at least one paid AI product have higher average revenue per user and net retention above 100%.

Twilio Chief Revenue Officer Thomas Wyatt described two AI-native customers that grew into $6 million and $9 million annual accounts within 14 to 18 months, adding messaging channels and software after starting with voice.

Analyze, Then Automate

RingCentral customer Destination Pet started with ACE to analyze call transcripts, then expanded to AIR Pro to automate appointment scheduling, confirmations and record updates. Makagon described a senior-living technology provider that added AIR Pro to verify callers, match them to records, identify the reason for the call and act on it.

"In just the first months, this customer achieved an over 95% accuracy rate in end-to-end workflow completion," Makagon said. RingCentral did not name the customer.

Faster Deployments

NiCE said Tripadvisor went from concept to live automated voice calls with NiCE Cognigy AI agents in two and a half months, and its AI agent posted a 90% customer sentiment score against 71% for human agents. Five9 CFO Bryan Lee said several backlog customers were ready to deploy AI faster than Five9 had forecast.

"That was primarily driven by several customers who are ready to deploy faster than what we forecasted," Lee said.

Cost and Experience Outcomes

Genesys reported several customer results this year. It said Electrolux achieved a 47-point increase in customer sentiment while cutting overall technology costs by 70%. It said Best Buy Canada reduced operating costs by 20% and Grupo Saesa cut handle time by 10% with AI virtual agents.

8x8 Chief Product Officer Hunter Middleton described the shift in terms many vendors used this season.

"AI adoption more than doubled, and it's showing up in live customer conversations, not pilots," Middleton said.

Genesys CEO Tony Bates said the company's results reflect "a broader shift from AI experimentation to enterprise-scale adoption."

Which Contact Center AI Use Cases Are Growing Fastest

Editor's note: The following table highlights the most important developments emerging from the contact center AI use cases customers are adopting fastest, from voice self-service to end-to-end workflow automation.

Use Case	How Customers Are Using It
Voice Self-Service	AI receptionists answer missed calls and capture leads that would otherwise go to voicemail. Small businesses lead adoption, and mid-market and enterprise accounts are starting to follow. Voice carries most AI customer interactions today.
Simple, Repetitive Requests	Customers start voice AI agents on routine jobs like password resets, then expand into complex, regulated scenarios. One bank resolves 70% of chat interactions with AI alone. Virtual agents are bought with a contact center platform or on their own.
Conversation Analytics and QM	Customers analyze call recordings and transcripts to identify why people call, grade agent performance and coach agents. Supervisors are adopting AI tools to monitor and manage teams.
Agent Assist	Customers pair virtual agents with AI-assisted human agents. AI is increasingly bought as part of contact center platform deals instead of as a separate product.
Lightweight Contact Centers	Businesses without a formal contact center buy bundled contact center products, and most of them add at least one AI product.
End-to-End Workflow Automation	AI agents verify callers, match them to records, identify the reason for the call and act on it. Customers automate appointment scheduling, confirmations and record updates, often after starting with analytics to find which calls to automate.