Buy, don't build: why Indian IT is on a $7 billion acquisition spree as AI eats its old model

India's IT services giants spent four decades perfecting a single machine: hire engineers by the hundred thousand, arbitrage global labor costs, and grow revenue roughly in line with headcount. Generative AI has broken the linkage — and the industry's response, visible in this year's deal flow, is to shop its way to a new model.

Indian IT firms are expected to spend $6.5–7 billion on acquisitions in calendar 2026, a record, as companies buy specialized capabilities rather than build them organically. Infosys has picked up healthcare technology firm Optimum Healthcare IT for $465 million and technology solutions provider Stratus for $95 million. HCLTech bought Singapore-based Finergic Solutions and paid $234 million for a 10.5% stake in sovereign-model startup Sarvam AI. Wipro acquired Mindsprint, the IT and digital services arm of Singapore's Olam Group, wrapped inside a multi-year transformation deal.

The deflation problem

The urgency is not hard to find in the earnings. AI-led productivity means clients expect the same outcomes for less: fewer billable hours, aggressive renegotiation at renewal, and pricing structures shifting from time-and-materials toward outcomes. Industry executives now openly discuss "AI-driven deflation" — the phenomenon of contracts shrinking 20–30% at renewal even as scope grows, because agents and copilots do work that juniors once did.

The counterweight is new AI revenue, and it is real but unevenly distributed. TCS reports annualized AI revenue crossing $2.3 billion; Infosys booked $275 million in a single quarter. But these figures remain slivers of $250-billion-plus industry revenue, and much of the "AI revenue" is classic integration work rebadged. The genuinely new capabilities — vertical AI products, proprietary platforms, regulated-industry depth — are precisely what the acquisition wave is trying to bolt on.

What the deals reveal

Read together, the purchases sketch the industry's theory of survival. Vertical depth over horizontal scale: Optimum (healthcare) and Finergic (financial services) buy regulated-domain expertise where AI deployment is hardest and margins are defensible. Products over projects: Mindsprint brings recurring platform revenue. And sovereignty as a market: HCLTech's Sarvam stake positions it for India's government-backed push for domestic models — a hedge that pure offshore delivery never needed.

There is also a defensive logic. The hyperscalers and Accenture are racing down-market with AI-native delivery, while a generation of AI-first startups — many founded by Indian IT alumni — sells "services-as-software" at a fraction of legacy pricing. Buying capability is partly about denying it to competitors.

The harder question

Acquisitions can buy technology; they cannot buy a new operating model. The core tension — 5 million-plus employees hired for a labor-leverage era, entering a software-leverage era — will not be resolved by tuck-in deals. The firms that navigate this will look less like outsourcers and more like product companies with services attached, running far fewer people per dollar of revenue.

Indian IT has re-invented itself before, from Y2K remediation to cloud migration. But those transitions expanded the demand for its people. This one is different in kind: for the first time, the industry is spending billions to acquire the very capabilities that make its traditional workforce less necessary. The $7 billion question is whether it is buying a future — or just buying time.