Nandan Nilekani’s AI warning for IT professionals: Four IT job roles vanishing but five positions are growing fast

Artificial intelligence is no longer just another tool for IT companies. It is beginning to change what technology professionals actually do every day. Speaking at Infosys’ Investor Day 2026, Infosys co-founder and chairman Nandan Nilekani said the shift is deep and structural, not cosmetic. According to him, writing code will not remain the central goal for software professionals as AI becomes embedded across systems and processes.

At the event, Nilekani said, “Talent will have to deal with the world where writing code will not be the goal, it will be actually making AI work.” He added that this transition will change both jobs and operating models.

He described the AI shift as a “root-and-branch” change. In his view, unlike earlier transitions such as mobile or cloud, AI cannot simply be added as a new layer. Companies will have to rethink customer journeys, internal processes, and even their organisational structures.


He also underlined that enterprises will need new capabilities in AI engineering, agent orchestration and handling non-deterministic systems, where the same prompt can generate different outcomes. This marks a departure from traditional software systems where outputs were predictable.

4 vanishing IT jobs

As demand pivots, some legacy roles are expected to shrink. Among the fastest declining jobs highlighted during the discussion were:

• Front-End Web Developers
• QA Testers
• IT Support Specialists
• Traditional blockchain-focused roles

These roles are not disappearing overnight, but the nature of work in these areas is changing as automation and AI tools take over repetitive tasks.

5 fastest growing IT roles

At the same time, several new and emerging roles are gaining ground:

• AI Engineers
• AI Forensic Analysts
• Forward Deployed Engineers
• AI Leads
• Data Annotators

Nilekani said, “Now I think talent transformation is huge. It's not that you will not need talent but it'll go from you know QA testing or development. We have all kinds of new roles AI engineers forward deployment engineers AI leads forensic analysts data.” He stressed that the real challenge for firms will be how to take their existing workforce and make sure they are reskilled and ready for new business models.

He was clear that there will still be a need for people, but “they'll be doing different things.”

Legacy systems remain a big hurdle

Nilekani pointed out that much of the productivity debate is focused on greenfield development. “Writing green field is not a big deal,” he said, explaining that generating large volumes of new code with AI tools is relatively easy. The real difficulty lies in modernising old systems.

He noted that companies have trillions of dollars invested in legacy systems, often with technical debt, data silos and undocumented dependencies. In some cases, organisations still rely on older experts to fix issues because no one else understands the system. According to him, modernising brownfield systems is far more complex than building something new from scratch.

AI adoption needs discipline

He also cautioned against superficial use of AI. “The very fact that you can generate stuff means you can generate slop,” he said, warning that without guidelines and quality controls, AI could create more confusion than value.

Nilekani explained that companies need usage guidelines, quality gates and explainability to ensure AI investments lead to real performance gains. Otherwise, what looks like productivity may not translate into measurable outcomes.

“The technology is far ahead of its deployment. Model performance is going up, but progress in implementing is not really there because implementing this is hard stuff. Fundamentally, it’s about organisational change, business change, retraining your people, changing your data so it’s no longer in silos,” he added.