Cognizant puts 18,000 engineers on Google Gemini for enterprise AI agents

Cognizant has assigned more than 18,000 of its employees to a new Gemini Frontier Capability Group — a dedicated unit built with Google Cloud to deploy AI agents for enterprise clients using Google’s Gemini models. The group launches globally this week, with no financial terms disclosed. It is not a joint venture, nor a co-branded product line: it’s internal capacity, scaled fast and focused on four priorities, modernising legacy systems, upgrading developer toolchains, automating business processes, and strengthening cyber defence operations.

All built atop Gemini, not third-party alternatives like Claude or Llama, though Cognizant maintains partnerships with OpenAI and others in parallel. The move signals how deeply Google is embedding Gemini into the enterprise delivery layer, not just as an API, but as the engine inside consulting-led transformations. Cognizant says more than 55,000 of its staff have already completed Google Cloud–led Gemini training, the largest such programme to date, per the company, and plans to roll Gemini access to 200,000 of its own employees by year-end.

That includes internal use of Gemini for code generation, documentation, and security analysis. The firm cited one unnamed retailer where Gemini-powered automation lifted developer productivity by 40% and cut related costs by 35%. Those figures remain unaudited and unverified by third parties. This isn’t Cognizant’s first Gemini play: it launched a smaller Gemini practice in April.