Infosys, TCS and the Nifty IT index fell even as SK Hynix profits surged and Anthropic filed for IPO. Here is the real reason Indian IT and the AI trade move in opposite directions.
On 27 and 28 October, something that looks completely backwards played out on Dalal Street. Global AI stocks were on fire, South Korea's SK Hynix reported a more than sevenfold jump in quarterly profit and announced it would return about $75 billion to shareholders, Anthropic filed its IPO paperwork, and the Kospi jumped 6%. You would expect Indian IT, a sector that lives and breathes technology spending, to join the party. Instead, Infosys fell 5.3% in two days, wiping out roughly Rs 25,000 crore in market value, and the Nifty IT index dropped over 3% in a single session.
This is not a one-off. It is a pattern that has repeated through 2026, Indian IT rallies when AI stocks abroad wobble, and sells off when they roar back. Understanding why requires looking at what Indian IT companies actually do for a living, and how that is different from what is driving the AI trade globally.
Rewind to mid-October. On 14 October, the Nasdaq had its worst day since July as investors dumped AI infrastructure names like CoreWeave and Oracle on worries about AI-linked debt. While Wall Street's AI darlings sold off, the Nifty IT index rallied nearly 11% over five sessions to cross 30,858, its best run in months. Capital that was fleeing expensive AI infrastructure stocks found a home in cheaper, cash-generating Indian IT names.
Then the mood flipped. By 27 October, Anthropic's IPO filing and blowout numbers from SK Hynix reignited enthusiasm for AI hardware. Money rotated straight back out of Indian IT and into semiconductor names in South Korea and Japan, Samsung Electronics, SK Hynix and Advantest all gained 4% to 9%. The Nifty IT index gave back a chunk of its October gains in just two sessions, with Mphasis and Infosys down 5% each, and Tech Mahindra, Coforge, TCS and HCLTech all down around 3%.
You can see the mirror image clearly when you place the two phases side by side.
| Period | What happened to global AI stocks | What happened to Nifty IT |
|---|---|---|
| 10 to 15 October | Nasdaq's worst day since July, AI infra names sold off | Up nearly 11% in five sessions |
| 26 to 28 October | SK Hynix profit up sevenfold, Anthropic files IPO, Kospi up 6% | Down over 3%, Infosys down 5.3% in two days |
The confusion comes from treating "AI" as one single theme. It is not. There are at least two very different trades hiding under that label, and Indian IT sits on the wrong side of one of them.
The first trade is AI infrastructure, chips, memory, servers and data centre capacity. This is what SK Hynix, Samsung, Nvidia and Anthropic represent. When this trade does well, it usually means hyperscalers like Microsoft, Google and Amazon are spending enormous sums on building AI capacity themselves, often in-house or through a small set of chip and cloud partners.
The second trade is AI services, companies that help ordinary businesses actually use AI, migrate systems, write custom software and manage IT operations. This is where Indian IT lives. And here is the catch, the more money hyperscalers pour into building AI infrastructure directly, the less certain the market becomes about whether that spending will flow down to traditional IT services vendors at all. Analysts have literally coined a term for this fear, the 'SaaSpocalypse', the idea that AI-native tools built by hyperscalers could eventually replace the layer of work that Indian IT companies have historically billed for.
So when AI infrastructure stocks rally hard, it is often read as a signal that capital and attention are consolidating around a handful of US chip and cloud giants, not spreading out to services vendors in Bengaluru and Pune. That is the mechanism behind the sell-off.
There is a second, more mechanical reason. Indian IT earns a large share of revenue in dollars, so currency moves matter directly to reported profit. We have covered this dynamic in detail in why IT stocks rise while the broader market falls, the weak rupee angle, where a weaker rupee against the dollar actually helps IT companies' reported earnings. The flip side applies too, when the rupee strengthens, as it did during the recent eight-day winning streak tied to Jio IPO inflows, it quietly works against IT margins even before you account for the AI rotation story.
Layer on top of that the fact that Infosys has already trimmed its full-year revenue growth outlook on softer volumes and pricing pressure, and you get a sector that was already fragile heading into this rotation, not one that suddenly broke because of a single day's AI headline.
If you track this blog regularly, you will recognise the pattern. We wrote about Nifty IT's sharp rally earlier in October when it became FPIs' favourite trade again, precisely because capital was avoiding expensive AI infrastructure names at the time. A similar swing happened back in December 2025 and January 2026, when the Nifty IT index saw a 'Golden Cross' formation and rallied 15% in three months, a move analysts directly attributed to capital rotating away from crowded US AI mega-caps toward broader, valuation-sensitive markets like India.
Earlier in the year, the opposite also played out, we covered why IT stocks slumped after Accenture's guidance cut and the subsequent 22% Accenture rally and whether to buy Indian IT before it. Global peer commentary, whether bullish or bearish, keeps setting the tone for how Indian IT trades, often more than domestic fundamentals do on any given day.
The practical takeaway is that Indian IT has effectively become a rotation play against global AI infrastructure sentiment, at least in the short term. That is uncomfortable if you are a long-term holder, because it means your stock can fall on days when the news looks objectively AI-positive for the world.
But the longer-term picture is more nuanced. If hyperscaler capex intensity normalises, as some brokerages expect once AI hardware productivity gains flatten out, services demand for migration, integration and operations work could re-accelerate, which is the scenario that drove the earlier Golden Cross rally. If instead AI infrastructure spending keeps compounding at the current pace, services vendors may keep playing second fiddle to chipmakers in terms of market attention, even if their own order books stay healthy.
For F&O traders, this kind of two-sided, headline-driven swing calls for defined-risk structures rather than one-directional bets. If you expect continued volatility in Nifty IT around AI headlines through earnings season, reviewing how straddles and strangles work can help you position for movement in either direction rather than guessing which way the next AI headline swings sentiment.
This inverse relationship will not hold forever, and it is not a law of markets, it is a reflection of where global capital happens to be focused right now. Keeping an eye on hyperscaler capex commentary and AI infrastructure earnings alongside Indian IT results, rather than looking at Indian IT in isolation, is probably the more useful habit for any investor tracking this sector through the rest of this earnings season.
Capital rotated into AI infrastructure names like SK Hynix and chipmakers after strong earnings and Anthropic's IPO filing, pulling money away from IT services stocks that are seen as a separate part of the AI story.
Not necessarily. The current weakness reflects short-term capital rotation toward AI hardware, not a confirmed structural decline, since IT services demand could re-accelerate if hyperscaler capex growth normalises.
It is market shorthand for the fear that AI tools built directly by hyperscalers could reduce the need for traditional software and IT services work that companies like Infosys and TCS have historically billed for.
Yes, since IT companies earn largely in dollars, a weaker rupee supports reported earnings, while a stronger rupee adds to margin pressure on top of any AI-related rotation.
It has repeated several times through the year, with Nifty IT rallying when global AI infrastructure stocks wobbled and falling when that AI trade strengthened again.