Cognizant has gathered more than 25,000 of its employees into a new OpenAI Frontier Capability Group, a practice set up to sell and deliver work built on OpenAI’s Codex and ChatGPT to large companies. The US technology services group announced it in a release on its newsroom on 30 October and in a post on its official X account at 10:04 GMT.
The number describes people Cognizant already employs, not new jobs. Chief executive Ravi Kumar S calls them “25,000 of our people”, and the release mentions no hiring. It also gives no financial terms, no revenue goal and no client names.
How the Frontier Capability Group is built
Cognizant says the group brings together professionals “aligned to delivering OpenAI products, including Codex and ChatGPT, for clients”. Its pitch is that access to frontier models is no longer the main barrier for many companies; the shortage is people who can build with those models inside the systems a business already runs.
The release says more than 21,000 Cognizant staff have earned over 40,000 OpenAI badges, each for completing a learning track on OpenAI technology. The group is expected to draw on what Cognizant calls its Frontier workforce model, which is developing “Frontier Certified Engineers and Frontier Business Operators credentialed directly by frontier-model companies”. Cognizant is also one of a small set of Advanced partners in the OpenAI Partner Network, and says its Codex hackathons in India and the Americas drew more than 12,000 participants.
Before offering the approach to clients, Cognizant says it tried it on itself. Its sales and customer success teams already use a tool built on OpenAI models to find and pursue new business. “We started with our own, and OpenAI is beginning to reshape how we sell,” Kumar said.
OpenAI’s chief revenue officer, Dali Rajic, is quoted in the release. “By bringing together more than 25,000 professionals in a dedicated OpenAI practice and putting our technology to work across its own operations, Cognizant is helping more enterprises turn AI into meaningful business impact,” he said.
Codex, old mainframe code and customer service come first
The group will start with three areas, according to both the release and the X post.
The first is what Cognizant calls AI-native software engineering, built around Codex, OpenAI’s coding agent, to help engineers write, test and ship code.
The second is application modernisation. Here Cognizant will use its own Cognizant Application Modernizer, powered by OpenAI models, from reverse engineering legacy mainframe code through to generating cloud-native code. The release adds that “engineers review every consequential decision”, and points to Cognizant’s experience modernising mainframe systems for more than 500 clients.
The third is agentic customer operations, applying OpenAI models to sales, service and support.
The one client result in the release is anonymous. Cognizant says it co-engineered an OpenAI-powered assistant for a global biopharmaceutical company that combines search, content creation, translation and approvals. It says the assistant has handled more than 5 million conversations and cut content creation time by 60%. Those figures are Cognizant’s own, and the client is not named.
The announcement comes a week after Cognizant launched AI Nexus, a five-part layer for metering and routing AI work across several model providers.
Cognizant’s OpenAI group is a staffing pledge with few outside checks
What Cognizant has announced is an internal reorganisation with a partner’s endorsement attached. The 25,000 figure measures how many existing employees it has pointed at OpenAI work, and the 40,000 badges measure training completed. Neither says how much client work the group has won or how many of those people are billing on OpenAI projects today.
Cognizant is not the only services firm to build a practice around one model company. HCLTech set up a dedicated Anthropic business unit this month, and software vendors such as Oracle and ServiceNow have widened their own OpenAI ties. Cognizant’s version stands out for its size and for the OpenAI quote, which shows the model company is happy to be associated with it, but no contract terms back it.
The most concrete parts are the human check in the modernisation work and the use inside Cognizant’s own sales teams, because both describe how the work is done rather than how many people are attached. The biopharma numbers are the only outcome offered, and with no client name or method behind the 60% figure, readers have only Cognizant’s word for it.
Cognizant says it has already modernised mainframe systems for more than 500 clients; the release does not say whether any of them have signed up with the Frontier Capability Group.
Sources
Primary:
– Cognizant: Cognizant commits more than 25,000 professionals to OpenAI delivery as enterprises move frontier AI into production (30 October 2026)
– Cognizant (@Cognizant) on X: OpenAI Frontier Capability Group announcement (30 October 2026, 10:04 GMT)
Press:
– None used for facts.
Related stories
- Cognizant Launches AI Nexus, a Five-Part Layer to Meter and Route AI
- Oracle and OpenAI Expand Pact With 4 Products, From ChatGPT to Codex
- ServiceNow Plugin Arrives in ChatGPT as OpenAI Partnership Widens
Follow @tsnmediahub on X for breaking AI news as it happens.
Follow on X





