Infosys boss describes AI job fears as ‘hogwash’

You have probably heard the argument that artificial intelligence will hollow out the middle of the workforce, automate coding and leave thousands of IT professionals scrambling for relevance. Nandan Nilekani, chairman of Infosys, calls that view "hogwash". Speaking at the Bloomberg New Economy Forum in New Delhi, he argued that AI will fuel demand for new human skills rather than eliminate the need for people. His reasoning turns on a distinction that many transformation programmes ignore: business change is not just about code.

Nilekani asked how organisations will ensure AI bots do not hallucinate, how they will understand customer needs, how they will collaborate across functions and get people to work together. The need for people is "more than ever before" for companies like Infosys, he said. He added that fears of AI-driven job losses are the result of "disaster marketing to suit the interests of large capital". Coding, he noted, accounts for only a fraction of any services project. The rest is context, judgement and the orchestration of work that no model can own.

What the sceptics say

Sam Kingston, CEO of software services company Zenitech and former UK head of EDS's private sector business, said Nilekani's argument "won't hold". He agreed to a point, then predicted a drop. AI will make people more efficient, speed delivery and reduce the number of senior and qualified people required to complete a project. Mid and junior level workers will be hit first as seniors do the work alongside AI. The traditional correlation between business growth and workforce growth at IT services firms will diminish. More business used to mean more people. AI is changing that.

Hexaware, a Tier 2 Indian firm with thousands of staff, offers a concrete example. Amrinder Singh, head of EMEA at Hexaware, told Computer Weekly last year that the company has a self-imposed target: by 2030 nearly 50 per cent of its work will be delivered, and revenue generated, through digital labour rather than human labour. Singh said that does not mean the human workforce will be smaller. With the growth of the industry fuelled by AI projects, more staff will be needed. But he admitted the growth of the human workforce may slow down.

Singh voiced a warning. There is no future for single-skilled people. Unless you are multi-skilled with domain understanding and an understanding of how to use AI and technology, you will not survive. Like workers in business roles, tech staff must retrain for a world with AI if they are to prosper.

The services model under pressure

Companies such as Infosys grew by offering IT skills to Western countries from India at a fraction of the cost of the same skills in Europe and the United States. Today those firms have established themselves on the top table when it comes to customer portfolios and the breadth of services they offer. While AI can now complete a significant chunk of coding requirements, Nilekani insists that accounts for just a fraction in any services project. The implication is that the margin pressure will land elsewhere: on the orchestration layer, the client-facing work and the domain expertise that turns a working prototype into a system people trust.

The tension between Nilekani's confidence and Kingston's caution reflects a deeper uncertainty about how AI will reshape professional services. Both agree that single-skilled workers face the greatest risk. Both acknowledge that AI will change the ratio of people to output. They differ on whether the new demand for skills will offset the efficiency gains quickly enough to protect employment levels in the near term.

What this means for L&D

Learning and development leaders in IT services and adjacent sectors should take three things from this exchange. First, the skills mix is shifting faster than headcount forecasts. Multi-skilling is no longer a nice-to-have. Domain understanding, AI fluency and the ability to collaborate across functions are now table stakes. Second, the middle tier of the workforce is most exposed. Junior staff who would have spent years building technical depth through repetition will need new pathways to mastery. Senior staff who can orchestrate AI-augmented delivery will be in demand, but only if they can work at a higher level of abstraction than the tools themselves. Third, the rhetoric around AI and jobs is polarised for a reason. Vendors, consultancies and capital all have incentives to frame the narrative in ways that serve their interests. The task for L&D is to cut through the noise and build capability around the work that remains stubbornly human: understanding what customers need, ensuring systems behave as intended and getting people to work together when the technology cannot.

Sources:

• Computer Weekly, 13 October 2026