Artificial intelligence (AI) represents one of the biggest opportunities the technology industry has ever seen but the challenge lies in execution, not invention, said Nandan Nilekani, chairman, Infosys, at the company’s “Investor AI Day” on Tuesday.
Speaking to analysts, Nilekani said there was “no opportunity gap” in the AI era.
Instead, what companies face is an implementation gap — a widening distance between the rapid advances in AI models and the ability of enterprises to deploy them effectively at scale.
“The opportunity is bigger than ever before. So don’t be distracted by that,” said Nilekani in his 20-minute address.
Citing Harvard University professor Clayton Christensen’s concept of “technology overshoot”, Nilekani noted AI capability might race ahead of enterprise needs, creating both risks and openings for those who could execute effectively.
Unlike earlier transitions — from mainframes to personal computers, or from on-premise systems to Cloud — AI is not merely a technological overlay. It requires fundamental changes in operating models, customer journeys, and business architecture.
“You cannot run businesses the old way,” he said. Legacy modernisation, long deferred by enterprises, can no longer be postponed. Many large firms still spend 60-80 per cent of their budgets of information technology maintaining aging systems built across decades. These systems, often in silos and undocumented, are also vulnerable to cyber threats.
To turn AI to good account, companies must undertake what Nilekani described as a “root-and-branch surgery” — cleaning up accumulated tech debt, breaking down data silos, and rearchitecting systems for AI-native operations.
A major implication of this transition is workforce transformation.
AI will not eliminate talent needs, Nilekani stressed, but it will redefine roles. Traditional coding and testing jobs will evolve into AI engineering, orchestration, forward-deployment engineering, forensic analysis and data-centric roles.
“You will need talent, but it’ll go from QA (quality assurance) testing or development. We have all kinds of new roles: AI engineers, forward-deployment engineers, AI leads, forensic analysts, data … so fundamentally, the challenge will be how you take your workforce and make sure that they are reskilled and ready for the new business? And that’s really the challenge that all the firms will face,” he added.
Infosys also for the first time gave details of its AI revenue.
Salil Parekh, chief executive officer (CEO) and managing director, in his presentation said the company provided AI service to 90 per cent of its 200 large clients.
Parekh said AI-related revenue was 5.5 per cent of the total in the third quarter of FY26.
For the October-December quarter of FY26, Infosys' consolidated revenue was Rs 45,479 crore, suggesting that its AI-related revenue was around Rs 2,500 crore for the said period.
“It is growing at a rapid pace. It is dynamic and working well with our clients,” said Parekh.
He highlighted six areas of growth in AI service. Based on an external analysis, he said these six areas had an opportunity in the range of $300 billion-400 billion till 2030. The areas are AI strategy and engineering, data for AI, process AI, agentic legacy modernisation, physical AI, and AI trust.
The company is expanding its partner ecosystem and also continues to recruit people from educational institutions, he added.
“We have recruited 20,000 graduates this year. For FY27 we will hire 20,000 graduates,” said Parekh.
Infosys on Tuesday announced collaboration with Anthorpic. Both will develop and deliver advanced enterprise AI solutions to companies across telecommunications, financial services, manufacturing, and software development.
Collaboration will focus on agentic AI by using tools like Claude Agent SDK. They will help clients build AI agents that can work across long, complex processes.
It will also help organisations modernise legacy systems, combining Infosys Topaz and Claude to accelerate migration and reduce the cost of updating aging infrastructure.
“There’s a big gap between an AI model that works in a demo and one that works in a regulated industry – and if you want to close that gap, you need domain expertise,” said Anthropic CEO and cofounder Dario Amodei.
Work will begin in telecommunications with an Anthropic Centre of Excellence to build and deploy AI agents tailored to industry-specific operations.
This will be done with Anthropic’s Claude Cowork and Infosys Topaz AI offers to help enterprises automate complex workflows, accelerate software delivery, and adopt AI with the governance and transparency that regulated industries require.
After the announcement, Infosys’s stock price rose 4.8 per cent and closed 1.9 per cent higher from the previous close.
 "AI offers huge scope but firms face widening implementation gap: Nilekani")
