Infosys CEO Salil Parekh disclosed at the company’s 45th Annual General Meeting that AI services revenue has reached approximately $1 billion in annualized run rate, according to HDFC Sky. Parekh signaled that AI agent modernization is now among the company’s largest service categories driving growth.
The shift is structural. Enterprise technology modernization has moved from traditional application development and cloud migration to what Infosys categorizes as agent-based technology modernization: deploying autonomous agents to automate enterprise workflows and integrate legacy systems.
What $1B From One Firm Signals
A $1 billion AI services run rate from a single IT services firm indicates that enterprise agent adoption has moved past pilot programs and into standard modernization budgets. Infosys is one of several major services firms (alongside Tata Consultancy Services, Wipro, Accenture, and others) competing for enterprise AI implementation work. If Infosys alone is generating $1 billion, the global enterprise AI services market is substantially larger.
The revenue structure also reveals how agentic AI scales beyond startups. Large IT services firms capture implementation value because they have existing enterprise relationships, security clearances, and domain expertise across industries. A startup may build the agent framework. Infosys sells the six-month engagement to deploy it inside a Fortune 500 company’s procurement system.
Timing and Context
The disclosure arrives at a moment when enterprises are simultaneously increasing AI budgets and imposing stricter governance on AI spending. Uber, Amazon, JPMorgan, and Meta have all implemented token-spending controls in recent weeks. The apparent contradiction resolves when you distinguish between uncontrolled tool adoption (which is being cut) and structured, project-based AI modernization (which is growing). Services firms like Infosys benefit from the latter: the more enterprises demand accountability and ROI from their AI spending, the more they need professional services to implement agents properly rather than letting employees experiment without oversight.