On October 13, 2026, at the Bloomberg New Economy Forum in New Delhi, Infosys co-founder and chairman Nandan Nilekani pushed back on forecasts that AI will hollow out Indian IT services. Business Today and an IANS report carried in the Free Press Journal both say his remarks followed IMF Managing Director Kristalina Georgieva's warnings about AI and jobs. Business Today quotes her description of AI as a "slow-moving tsunami" for labour markets. The IANS report says she warned AI could affect nearly 40% of jobs worldwide.
His argument
Nilekani's case, as Business Today quotes it: "Requirement for companies like Infosys is increasing. We are going to see more work coming our way. Predicting doom for such companies is just disaster marketing." His reasoning is that adopting AI is a business transformation project, not a coding project, and that "technology is only 30 per cent in technology project". The rest, he said, is governance and guardrails. On where the jobs will come from: "New jobs will be in understanding client requirements." He also pointed to small businesses: "Once we unleash small business economy, a lot many jobs will be created."
He also argued that AI will not stay concentrated in a few companies. "You cannot bottle this technology," he said, as more chips and compute let more players build and deploy AI. Business Today sums up his view for India: "India has an opportunity to lead, but not by building the most powerful AI models." The IANS report says he sees India's advantage in its experience deploying digital infrastructure to very large populations.
The other side of the same week
One day earlier, HCLTech's CEO C Vijayakumar gave a more cautious view of the same market. He said AI-driven productivity is already lowering prices in core services contracts, and that "we are in the early phases of deflation." The two views can both be true. Nilekani is describing more total work, much of it outside coding. Vijayakumar is describing lower prices for the coding and operations work these firms already do.
What remains uncertain. Nilekani gave no figures for Infosys's own AI demand or headcount, and Infosys's next quarterly results will be the first test of the claim. He also chairs the company whose business model is under discussion, so his view is not a neutral forecast.
Analysis: His 30% figure is the most useful claim for engineering leaders, and the easiest to test. If most of the cost of an AI rollout is requirements, governance, guardrails and change management, then cheaper code generation reduces only part of the cost. The remaining work moves to people who can specify what the system should do and check that it does it. For Indian services firms that suggests training for requirements and assurance roles, not only for coding with agents. For buyers, it means pricing should be negotiated separately for the coding work, which AI is making cheaper, and the governance work, which it is not.
At the Bloomberg New Economy Forum in New Delhi on October 13, Nandan Nilekani called predictions of doom for IT services 'disaster marketing', said technology is only about 30% of a transformation project, and argued India should lead on scaling AI to many users rather than on building the most powerful models.