Cognizant said on 10 November that it had formed a dedicated Claude delivery group with more than 25,000 professionals and was introducing a service to run Claude-powered systems for clients. The Nasdaq-listed IT services company, based in Teaneck, New Jersey, set out the plans in a company announcement distributed by PR Newswire. The proposed service would put Cognizant in charge of operating systems built with Anthropic’s Claude models after they enter production.
Key points
- Cognizant says its Claude Frontier Capability Group brings together more than 25,000 professionals.
- The company says more than 44,000 associates are trained or certified on Claude.
- The managed service combines Cognizant staff, licensed Claude access and the company’s own software.
- Cognizant announced a separate OpenAI delivery group of more than 25,000 professionals on 30 October.
Claude group draws on Cognizant’s trained workforce
The Claude Frontier Capability Group is intended to build and operate systems using Anthropic’s AI assistant inside client businesses. Cognizant says more than 44,000 associates across its business units and corporate functions have already been trained or certified on Claude. The company describes the new group as a dedicated global organisation for client delivery, rather than a count of everyone with Claude training.
Cognizant holds more than 26,000 Claude certifications and ranks first globally for Claude Certified Architect: Foundations, Business Standard reported. Certifications and trained associates are separate measures: Cognizant’s workforce figure includes people who are trained or certified, while the certification figure counts credentials.
The group will initially work on redesigning business domains around AI, modernising the systems that support them, operating AI agents securely at scale and making AI-native software engineering part of delivery, Business Standard reported. Cognizant announced in November 2025 that it was deploying Claude across its own workforce, including corporate functions and engineering and delivery teams.
The Claude initiative also draws on Cognizant’s Frontier workforce model. Under that programme, the company says it is developing Frontier Certified Engineers and Frontier Business Operators who receive credentials directly from model companies. Ravi Kumar S, Cognizant’s chief executive, said clients want a provider to take responsibility once AI systems are operating within their businesses.
Cognizant offers to operate Claude in production
Under the managed services model, Cognizant would operate clients’ Claude-powered systems in production. Its proposed package combines dedicated personnel and licensed access to Claude with Cognizant intellectual property, including its AI Delivery Operating System. The company says it would tailor security and governance controls to each client’s connectivity, architecture and risk requirements.
Cognizant says the arrangement is designed to give clients more predictable AI spending and one partner responsible for running the systems. The operational responsibility described in the announcement extends beyond building a Claude application: Cognizant would remain involved when that application is used in the client’s business.
“We are putting more than 25,000 of our people on Claude and taking on the job of running it for clients through managed services,” Kumar said in the announcement. Paul Smith, Anthropic’s chief commercial officer, described finding people able to make AI work inside a business as a challenge for enterprises. Anthropic builds the Claude models that Cognizant plans to deliver and operate for those clients.
OpenAI group offers a delivery comparison
Cognizant announced an OpenAI Frontier Capability Group with more than 25,000 professionals on 30 October, according to its earlier release. That group is aligned with OpenAI products including Codex and ChatGPT. Cognizant said it would initially concentrate on software engineering, application modernisation and customer operations.
The OpenAI release described a different route into client work. Cognizant said its sales and customer-success teams already used an OpenAI-based system to find and pursue opportunities, giving the group an internal deployment to draw on. It also reported more than 21,000 associates had earned over 40,000 OpenAI badges for completing learning tracks.
For a client example, Cognizant said it had co-engineered an OpenAI-powered assistant for a global biopharmaceutical company that handles search, content creation, translation and approvals. The assistant had processed more than 5 million conversations and reduced content-creation time by 60%, according to Cognizant’s 30 October release.
