AI growing, IT services isn’t yet – Opinion News

The IT services industry has spent two years explaining why AI will create more work than it destroys. This quarter, the numbers finally make that argument look plausible. They also make clear that the benefits will be distributed very unevenly.

Start with the broad picture. TCS, HCLTech, and Wipro have now reported their September quarters. Persistent has too. Infosys, unusually for this comparison, reports its September quarter only on October 23, so its June quarter is the latest available result. Accenture, meanwhile, has reported its August year-end and, more importantly, laid out a three-year growth ambition that tells us where the industry leader thinks the market is going.

The aggregate picture is modest growth, respectable margins, healthy deal pipelines, and increasingly material AI revenue. But averages conceal more than they reveal.

TCS reported $7.64 billion of quarterly revenue, up only 0.5% sequentially in constant currency. Yet its annualised AI revenue crossed $3.1 billion, more than 10% of revenue, while quarterly contract value was $9.6 billion. Operating margin remained a formidable 24%. The company is not growing quickly, but AI has moved from presentation slides into a business of genuine scale.

HCLTech looked stronger. Constant-currency revenue rose 2.3% sequentially and 2.6% year-on-year, while new deal wins reached $3.26 billion. Its advanced AI revenue rose 78% year-on-year to $219 million. HCL also raised its full-year revenue growth guidance to 2.5-4%. Perhaps most interestingly, after reducing employment in the previous quarter, it added 6,450 people. AI productivity and hiring, apparently, can coexist.

Then there is Wipro. IT services revenue was essentially flat sequentially and up just 0.6% year-on-year in constant currency. Total bookings fell 22.9% sequentially and large-deal bookings collapsed 58.6% to $674 million. Its December-quarter guidance, minus 1.5% to plus 0.5%, suggests that the industry’s rising AI tide is not lifting every boat. Wipro is the clearest large-company outlier.

Persistent is the opposite outlier. Revenue rose 7.8% sequentially and 20.8% year-on-year in constant currency, its 26th consecutive quarter of sequential growth. At $488 million quarterly revenue it is much smaller than the majors, but that is precisely the point. Its concentration in digital engineering, cloud, and AI-heavy work puts it closer to where discretionary technology spending is moving. Its acquisition of Nagarro also signals an ambition to convert that positioning into global scale.

Infosys’s latest available quarter showed the middle path: 2.4% year-on-year constant-currency growth, 21.1% operating margin, $3.6 billion of large deals, and AI revenue already at 8.2% of total revenue. But it narrowed full-year growth guidance to 1.5-3%. Again, AI is growing much faster than the company containing it.

That contradiction may be the most useful way to understand the sector. AI revenues are expanding rapidly; company revenues are not. The new business is partly replacing, compressing, or repricing the old business rather than simply sitting on top of it.

Which brings us to Accenture.

Its fiscal fourth quarter produced $18.7 billion of revenue, growing 7% in local currency, with $22.2 billion of bookings. For fiscal 2026, revenue reached $74.2 billion and bookings $84.5 billion. But the more consequential announcement came at its investor day last week. Accenture expects mid-single-digit revenue growth in fiscal 2027 and 2028, stepping up to mid-to-high single digits in 2029, while expanding operating margins by 10 to 30 basis points annually.

That sounds conservative until one examines how it intends to get there.

Accenture has expanded its addressable market estimate to $3 trillion. It plans to keep winning the giant enterprise transformations that built the firm, while pushing into the mid-market through Accenture Edge. Its target segment there alone represents an estimated $240-billion market. It is simultaneously expanding AI and data talent, building specialist capabilities, acquiring firms, and turning what it calls “reinvention” into the organising principle for consulting and managed services. The important phrase from management is that it intends to grow while “absorbing AI efficiencies”.

That is a remarkably concise description of the industry’s next problem. If AI allows the same project to be delivered with fewer people and fewer hours, providers have to give some productivity benefit back to clients. Growth therefore requires winning more work, entering new markets, moving further into business processes, and selling outcomes rather than labour. Accenture is explicitly designing itself around that reality.

Indian IT companies are heading in the same direction, but at different speeds. TCS has scale, margins, and rapidly growing AI revenue. HCLTech currently has momentum. Persistent has growth. Infosys has strong economics but subdued guidance. Wipro still has to demonstrate that its strategy can translate into sustained revenue.

The earnings therefore contain both reassurance and warning. Enterprise technology spending has not disappeared. AI is creating large new pools of work around data, cloud, modernisation, security, and process redesign. But it is simultaneously attacking the industry’s old unit of sale: the engineer-hour.

For 30 years, Indian IT mastered the economics of putting more people against more technology work. The winners of the next decade will master something harder: producing much more value with fewer people, while somehow persuading clients to pay for the value rather than the people.

That is why these results look less like the beginning of a spending recovery than the opening of a competitive cycle. Demand is returning selectively, but the basis of competition is changing underneath it. Revenue growth now depends on converting productivity into demand before productivity erodes pricing.

The writer is technology consultant and venture capitalist by invitation

Disclaimer: The views expressed are the author’s own and do not reflect the official policy or position of Financial Express.