Nandan Nilekani: AI Is Driving the Biggest Overhaul of IT Services in Decades

Infosys co-founder Nandan Nilekani told Bloomberg that AI is launching the largest transformation of the IT services industry in decades. By his assessment, the company, founded in 1981, gets new opportunities from this, not just risks. The conversation with Shery Ahn took place at NEF in New Delhi. Nilekani talked about AI adoption in large corporations, legacy systems, and Infosys's model-agnostic approach.

The published announcement contains no revenue figures, forecasts, or contract names. What remains is the position of a man who built Indian outsourcing from scratch, and the logic behind that position.

Why IT services take the hit

The classic model of the Indian IT giants rests on large teams of engineers. They write, test, and maintain code for clients around the world. Generative models noticeably change exactly these tasks: code generation, test writing, documentation, first-line support. So the phrase "biggest transformation in decades" refers not to an abstract industry but to Nilekani's own business model.

He chooses the framing of opportunity. If models do the routine work, an integrator can sell outcomes rather than person-hours. The open questions are how quickly clients will agree to pay for outcomes and who will capture the saved margin.

Legacy as the main barrier and the main source of work

Nilekani discussed legacy systems separately. It is a precise hit on the problem. Banks, insurers, retailers, and government agencies have lived for years on old platforms with fragmented data. The best model on the market is useless if it cannot be safely connected to that data and those processes.

This is where the integrator has an advantage. It knows the client's tangled infrastructure, has people for migration, and has the customer's trust. Labs like OpenAI or Anthropic sell models and APIs, but they have no time to plow through a client's decade-old technical debt.

A bet on model independence

According to Nilekani, Infosys follows a model-agnostic approach. The company does not tie itself to a single provider and instead picks a model based on the task, the budget, and the client's data requirements. The logic is pragmatic. Models get cheaper quickly and trade places in the rankings every few months. Lock-in to one vendor is risky for both the integrator and the client. Value shifts to the layer of integration, orchestration, and quality control.

This strategy has a weak spot. Cloud platforms and the labs themselves are increasingly going directly to enterprise customers, with agents and ready-made solutions. Neutrality pays off only as long as the client needs an independent advisor rather than a single turnkey platform.

Key takeaways from the interview:

• AI is called the biggest transformation of IT services in decades.
• For Infosys, it is a source of new opportunities, not just pressure on revenue.
• Adoption runs into clients' legacy systems.
• The company works with multiple models and is not tied to a single provider.
• Venue: NEF, New Delhi; the interview was conducted by Shery Ahn (Bloomberg).

What clients and investors should watch

For now, the statements remain positioning. They are worth checking against concrete indicators. First: the share of projects where payment is tied to results rather than hours. Second: revenue growth against the backdrop of code automation, that is, whether new orders offset the lost routine work. Third: how the company handles data security when working with multiple models.

If you are a client, do not ask your contractor "which model do you use"; ask how it will swap models without rewriting integrations. If you are an investor, follow management's quarterly comments on pricing and headcount. That is where it shows whether Nilekani's optimism is justified.