Infosys (NSE/BSE/NYSE: INFY) reported a quarter that pulls in two directions. Revenue for the three months to 30 September 2026 was $5,145 million, up 1.5% year on year and 1.2% sequentially in constant currency; operating margin held at 21.0%; and free cash flow was $928 million. AI revenue reached 10.5% of the total, up from 8.2% in the prior quarter and growing 29% sequentially. Large-deal total contract value was $5.2 billion, including a net-new mega deal of more than $2 billion. And yet the company cut its full-year revenue-growth guidance to 1–2% in constant currency, from 1.5–3%, and its US-listed shares fell sharply.
The quarter matters because Infosys is the industry’s second-largest firm and a bellwether for how enterprise technology spending is shifting. When a company of its scale cuts its growth forecast while reporting fast-growing AI work, that combination is not a contradiction to be resolved but a signal about the structure of the market: the money is moving toward AI, and some of it is moving away from the traditional services that AI makes cheaper. Both facts matter, and together they describe the moment Indian IT services is in: AI is producing fast-growing revenue while simultaneously compressing the business around it.
The AI revenue milestone
By Infosys’s own measure, AI is now a meaningful slice of its business. AI revenue crossed 10% of the top line for the first time, having stood at 8.2% a quarter earlier — implying roughly $539 million in the quarter. Chief executive Salil Parekh said the growth spanned all six areas of the company’s AI strategy: AI strategy and engineering, data for AI, process AI, agentic legacy modernisation, physical AI and AI trust. “As enterprises adopt AI at scale, another quarter of strong growth in AI revenues is a recognition of our differentiated market positioning and client confidence,” he said.
Infosys is not alone in reaching the milestone. In the same quarter, TCS reported annualised AI revenue of about $3.1 billion, up 19.2% sequentially, and said AI margins were above its average; HCLTech reported advanced-AI revenue of $876 million on an annualised basis, up 72% year on year. For the first time, AI contributed more than 10% of revenue at two of India’s largest IT firms.
The disclosure caveat
That milestone deserves a health warning, because “AI revenue” is not a standard measure. Infosys publishes a percentage with no definition and no dollar figure in its release; independent analysis has found no consistent methodology across the industry. Accenture’s final AI metric explicitly excluded data work, classical machine learning and robotic process automation; HCLTech counts a narrower “advanced AI” subset; TCS annualises a single quarter’s AI revenue. The disclosed shares across firms that report them range from roughly 4% to 10%, and they cannot be added up cleanly.
There is a precedent for treating such metrics cautiously. Infosys reported a “digital” revenue share from 2018, which climbed from about 25% to roughly 63% before the company retired it in 2023, once it covered so much of the business that it stopped distinguishing anything. AI revenue may follow a similar arc — which is why the number is best read as directional, not as a like-for-like comparison.
The guidance cut and AI deflation
The guidance cut is the more consequential disclosure, and AI is part of the reason. Infosys said growth was below its expectations, with softer volumes and lower-than-anticipated price increases; revenue from North America, more than half its sales, fell 1.5% — its weakest showing in two years. “We have also seen pricing pressure on the back of competitive intensity and higher ask on productivity,” chief financial officer Jayesh Sanghrajka said on the earnings call, adding that the pressure would continue into the second half alongside higher compensation costs.
That is AI deflation in practice. AI lets providers deliver work faster and with fewer people, but clients increasingly want those savings passed on as lower prices, smaller teams or outcome-based contracts. Everest Group’s Peter Bendor-Samuel estimates that deflation in IT services, historically around 2%, is now running at about 5%, while industry growth has slowed from roughly 6% to 2.2%. Forrester’s Biswajeet Mahapatra frames it as a trade-off: AI improves near-term delivery economics, but the longer-term effect is deflationary for revenue as the link between headcount, effort and billing weakens.
The dynamic is not unique to Infosys. Across the sector, firms are expanding operating margins on AI-led productivity while warning that the same productivity is deflationary for revenue. The apparent paradox resolves once you separate cost from price: AI lowers the cost of delivering work, and competition passes much of that saving to clients as lower prices, so margins can improve even as top-line growth stalls. That is a harder business to grow than a labour-arbitrage model, and it is the one every large services firm now has to run. Infosys’s own numbers are the illustration — AI revenue up 29% sequentially, total growth at the weakest pace in six quarters.
The deal book and the margin
The offsetting positive is the deal book. Infosys won 19 large deals worth $5.2 billion, its second-highest quarterly total in five years, 62% of it net new. That included two mega deals: a net-new engagement of more than $2 billion over seven years with a financial-services client, and a $1 billion renewal. The catch is timing — the net-new mega deal begins transitioning only in the second half of the fourth quarter, so it adds very little to FY27 revenue and is expected to contribute more than one percentage point to growth in FY28 instead.
Margin held at 21.0%, the midpoint of the guided range, despite the pressures. Infosys credited its internal cost programme, Project Maximus, and currency tailwinds for offsetting headwinds including post-sales support costs and acquisition amortisation. The composition of the deal book reinforces the point. Large deals in this market skew toward cost takeout and vendor consolidation — work that is won on price and executed with fewer people — rather than open-ended transformation programmes that scale headcount. A $2 billion mega deal is a vote of confidence, but its economics are built on efficiency, which is exactly the force weighing on the sector’s growth. But the company flagged compensation increases and continued pricing intensity as headwinds for the second half.
What to watch
Three things will show how the story develops. First, whether Infosys publishes a definition and dollar figure for AI revenue, which would let investors judge how much is genuinely new work rather than reclassified services. Second, whether the net-new mega deal ramps as forecast in FY28, since its contribution is the company’s main answer to the weak growth outlook. Third, whether AI deflation eases or deepens — the variable that decides whether AI revenue growth translates into overall growth or continues to cannibalise it.
The macro picture adds to the uncertainty. Client spending remains cautious amid geopolitical tension, inflation and shifting technology priorities, and discretionary spend — the higher-margin work that lifts growth — has been slower than usual. Infosys said sectors including financial services, energy and utilities and manufacturing were relatively strong, while retail and hi-tech were weak, with hi-tech hit by lower third-party revenue and reduced spending by a few large clients. Geography matters too: North America is the weak spot, and Infosys said it could remain a laggard for several quarters.
For investors, the practical question is whether AI becomes a net growth engine or a permanent drag. The bull case is that AI opens a large new market — Infosys has put the opportunity at $300 billion — and that early leaders will capture share. The bear case is that AI work, however fast it grows, forces the company to pass savings on to clients, so a larger AI share may not translate into company-wide growth. Infosys’s quarter, with AI rising and guidance falling, is evidence for both. The coming quarters will show whether the guidance cut marks a trough or the start of a longer slowdown, and whether the AI opportunity proves additive or substitutional for the industry’s revenue base.
What the results do and don’t establish
They establish a quarter of strong AI-revenue growth, a record large-deal count, a stable margin and a guidance cut. They do not establish a standard definition of AI revenue, how much of it is incremental rather than reclassified, or that AI-led productivity will translate into faster overall growth rather than lower prices.
The honest summary: Infosys has built a genuine AI franchise, but it faces a difficult growth outlook, and the same factors shape both. AI is now a tenth of the business and growing fast, but it is also the force letting clients demand more for less, which is why the guidance came down even as the AI number went up. The milestone is real; the metric behind it is soft; and the quarter’s real lesson is that for IT services, AI is not only a new revenue line but a deflationary pressure on the old one.
Editor’s Note: This article draws on the Infosys release and the company’s investor materials and earnings-call transcript, plus independent coverage by Reuters (via StreetInsider), CNBC-TV18, Moneycontrol, Business Standard and Investing.com, and AI-disclosure tracking analysis. The company’s filing provides the financial figures. The company defines its AI-revenue percentage but discloses neither its definition nor the dollar amount. Peer companies use different definitions for their AI figures. No interview took place.