Cognizant is the most aggressive bet in the group: charging for what the AI does, not just for the hours people work. Under CEO Ravi Kumar it frames the future as a four-step climb, from delivering projects, to billable hours, to owning outcomes, and finally to underwriting and being paid for those outcomes. Its agent stack runs on Neuro and the Agent Foundry.

That means rebuilding the delivery model. Project Leap reshapes the staffing pyramid into something broader and shorter, hiring more new graduates and using AI agents so people reach expertise faster, and Cognizant markets an AI builder strategy aimed at closing what it calls the AI velocity gap between what clients spend on AI and what they actually get from it.

The boldest claim, getting paid for outcomes instead of effort, is also the riskiest. Cognizant has openly admitted AI deflates the old billable-hours business, and it has not yet disclosed a standalone AI run-rate, so the question is whether outcome-priced revenue grows faster than its traditional hours shrink.

The upshot: no peer has leaned harder into being paid for the AI work, and management is backing that conviction with capital, but the economics still have to prove the new model outruns the one it is cannibalizing.