A story about the strange new world where firing people is actually a sign of health.
Here's something that sounds like a riddle.
TCS — India's largest IT company — cut 23,000 employees in FY26. Revenue went up. Margins held. And the stock, after a rough patch, still attracts "buy" ratings from major brokerages.
How do you cut tens of thousands of people and grow at the same time?
Welcome to the AI transformation of Indian IT. And it's stranger, and more consequential, than most people realize.
First, what's actually happening?
TCS CEO K Krithivasan clarified that the 23,000 job cuts in FY26 were to let go of senior and middle managers who couldn't adapt to new work models, not direct AI automation. These weren't coders being replaced by ChatGPT. These were layer-upon-layer of project managers, account managers, and bench-sitters — the structural fat of a company that hired aggressively during the 2021–22 tech boom.
But here's the honest part: AI is changing what a 100-person software team can deliver. Infosys helped Hertz migrate 3 million lines of COBOL code to a modern microservices environment using AI foundation models — at 60% lower cost and 60% faster timeline compared to how it would have been done without AI.
If you can do a job at 60% lower cost and 60% faster, you need fewer people for the same revenue. That's the AI productivity math — and it's compressing IT headcount even as deal values grow.
The revenue picture — surprising resilience
TCS reported annualized AI revenue over $2.3 billion in Q4 FY26, showing early earnings from these new AI capabilities. Infosys CEO Salil Parekh has confirmed that AI revenue is "growing very nicely" and crossed 5.5% of revenue in Q3 FY26 — and is higher now, though the company won't disclose the exact number.
Infosys has strategic AI collaborations with Google Gemini, NVIDIA, Microsoft, AWS, and Anthropic — and has deployed over 30,000 developers on GitHub Copilot.
This is the shift: Indian IT companies are no longer just staffing agencies that provide offshore programmers. They're becoming AI transformation partners — helping global corporations redesign their operations using AI agents.
The rupee bonus — a silent multiplier
Here's the part most investors miss.
The rupee has fallen from ₹85 to ₹95 against the dollar over the last year. For a company like TCS, which earns roughly 95% of its revenue in US dollars and pays most of its costs in rupees, this is a massive tailwind.
Every dollar of revenue now converts to ₹95 instead of ₹85. On TCS's revenue of ~$30 billion, a ₹10 depreciation adds roughly ₹30,000 crore to reported INR revenues — without a single extra client or contract.
For Infosys (revenue ~$18 billion), it's a ₹18,000 crore tailwind. For HCLTech (~$14 billion), roughly ₹14,000 crore. All without selling a single extra line of code.
But what about the worry — that AI will eat IT revenues entirely?
HCLTech's stock dropped 9.7% on April 22, 2026, after a 3.3% revenue dip and cautious FY27 guidance. The Nifty IT index has seen a ~25% correction year-to-date as of March 2026, largely due to fears of AI causing lower revenues through automation.
The market's fear is this: if AI can do coding work 60% cheaper, will clients pay Indian IT companies 60% less? That's the bear case.
The bull case — articulated by TCS CEO Krithivasan: "We may need less people for doing what you're doing today, but you're going to be doing much more." The size of the work is expanding. AI makes it possible to take on more projects, not fewer.
Which view is right? Probably both — for different segments. Routine coding will compress. AI consulting, agent deployment, legacy modernization — these are growing. The companies that navigate this transition win. The ones that don't will see the AI era not as an opportunity, but as an existential threat.
The Indian IT sector is in the middle of its most consequential transformation since it was born. And the next 3 years will determine whether it emerges stronger — or smaller.
