India’s IT selloff is no longer just a valuation scare. Investors are starting to price in the possibility that AI will shrink the old outsourcing model before it creates a larger new one.
Tata Consultancy Services did not fall 9% on June 3 because one bad trading session suddenly changed the company’s future. It fell because the market is losing patience with a business model built around large teams, long projects and steady billing at the exact moment AI tools are making software work faster, cheaper and less dependent on headcount.
According to Reuters, India’s Nifty IT index was down 5.8% at 29,310.25 points on Wednesday, putting it on track for its worst day since February 4. TCS led the decline, while Infosys dropped 4.3% and Wipro fell 3.7%. Coforge and Persistent Systems also lost 5.7% each, showing that the pressure was not confined to one heavyweight stock.
That matters because India’s IT services sector is not a niche corner of the market. It is a roughly $300 billion industry, a major export engine and one of the country’s most important white-collar employers. For years, its strongest listed companies traded like reliable compounders. Now investors are asking a harder question: what happens when the work clients used to outsource can be automated, compressed or handled by smaller internal teams using AI agents?
The classic Indian IT formula was simple and powerful. Hire skilled engineers at scale, deliver software and support work for global clients, and turn wage arbitrage into dependable margins. TCS, Infosys and Wipro became global names by executing that model better than almost anyone else.
Generative AI does not destroy that model overnight, but it weakens the assumptions behind it. Coding assistants, automated testing tools, AI support agents and enterprise workflow systems can shorten projects and reduce the number of people needed to deliver the same outcome. For clients, that is the point. For services companies paid partly through effort, time and staffing intensity, it is a direct challenge.
This is why the selloff followed a brief rally rather than a fresh earnings shock. The Nifty IT index had gained about 7% over the previous two sessions as investors hoped rising AI spending would support demand for implementation work. Wednesday’s reversal showed the other side of that trade. AI may create new projects, but it may also lower the price of old ones faster than new revenue arrives.
Kotak Institutional Equities has estimated a net revenue deflation impact of about 3.5% annually from generative AI adoption over FY2027 to FY2029, after allowing for some offset from fresh AI-related work. That is not catastrophic in isolation. But for an industry already dealing with cautious client spending and low single-digit guidance, it changes the earnings math quickly.
AI demand is real, but timing is the problem
The bullish case is still credible. Large enterprises need help modernising legacy systems, cleaning up data, building AI governance, securing models and integrating agents into workflows that actually run a bank, insurer, retailer or manufacturer. Those are not weekend projects. They require domain knowledge, engineering discipline and trust, all areas where Indian IT majors have deep client relationships.
The problem is timing. AI budgets are moving first toward chips, cloud capacity, foundation model licences and software platforms. Gartner’s forecasts cited in recent market coverage point to global technology spending growing 13.4% in calendar 2026, while IT services growth is expected at only 4.2%. In plain terms, the pie is growing, but the slice Indian outsourcers traditionally eat is not growing at the same pace.
ICICI Direct’s sector analysis puts the risk in sharper terms. It says nearly 30% of the IT services industry is exposed to GenAI-led deflation, with a base-case impact of about $40 billion for India and a worst-case impact of $80 billion to $85 billion in the at-risk segment. Those figures are estimates, not destiny, but they explain why investors are no longer treating AI as only a consulting opportunity.
The companies know this. Infosys has pushed its Topaz platform, TCS has spoken about AI-led services and Wipro has been trying to position Holos as part of its enterprise AI pitch. The challenge is that investors want evidence that these platforms change delivery economics, not just branding. Selling AI services wrapped around third-party models is useful. Rebuilding the operating model around AI-native delivery is harder.
That is the real test for management teams. If AI makes an engineer 30% more productive, clients will expect some of that gain through lower prices, faster delivery or broader scope. Vendors can protect margins only if they redesign teams, contracts and tools quickly enough to keep a fair share of the productivity benefit. Otherwise, AI becomes a deflation machine controlled by the customer.
The market is not saying TCS, Infosys and Wipro are finished. It is saying the easy premium attached to their old growth model needs to be earned again. Watch deal renewals, revenue per employee, margins on AI-heavy projects and whether headcount growth decouples from revenue growth. Those indicators will tell us whether this selloff was panic, or the beginning of a more serious repricing of Indian IT.
Also read: Suno's new valuation puts AI music's legal gamble in plain view • Broadcom faces an earnings test after a huge AI stock rally • AI worms have moved from theory to demonstrated cyber risk
