Will AI Create More Jobs Than It Destroys in India's Services Sector?

There's a version of this question that gets asked at every tech conference, in every LinkedIn post, in every panel discussion with a moderator who opens with "so, should we be worried?" And there's a version that actually matters to the twenty-four-year-old sitting in a Bengaluru WeWork right now, three years into a BPO job, watching her manager demo a new AI tool that does in ten minutes what used to take her team a day.

Those two versions of the question don't have the same answer.

The comfortable macro story

Zoom out far enough, and the numbers look fine. Nasscom's 2026 Strategic Review pegs the Indian tech industry at $315 billion in FY26, growing 6.1% year on year. The sector added a projected 135,000 net jobs this year, pushing total headcount toward 5.95 million. AI-related roles are multiplying too — demand for AI-skilled professionals is expected to cross a million by 2026, and postings for Gen AI developers have grown sharply since 2022.

Read only those lines and you'd write an optimistic headline. AI is creating a new category of work, and India, with its scale and cost advantage, is positioned to capture a big share of it.

But that 2.3% hiring growth rate is the tell. A tech industry growing revenue at 6% while hiring at 2.3% isn't growing jobs and productivity at the same pace — it's growing output per employee, which is a polite way of saying each employee is now expected to do more. That's not a crisis by itself. It's what productivity gains have always looked like. But it does mean the sunny macro numbers are hiding something the micro numbers show more honestly.

The jobs that go first aren't the ones people worry about

The popular fear is that AI comes for senior, high-value roles — the architects, the strategists, the people with twenty years of judgment built up. In practice, it's going the other way. The tasks that AI automates first are the ones that used to be how junior people earned judgment: note-taking, first drafts, basic research, routine QA, entry-level coding, tier-one support tickets. Nasscom's own analysis on the services sector flags exactly this — the rungs at the bottom of the career ladder are the ones AI is quietly removing, which raises an uncomfortable question about where the next generation of experienced professionals is actually going to come from if the entry-level jobs that built the current generation no longer exist in the same volume.

This is the part of the "jobs created vs. jobs destroyed" debate that a simple net-number obscures. Even if the totals balance out, you can still end up with a labor market that's hollowed out in the middle — fewer entry points, fewer people getting the on-the-job reps that used to turn a fresher into someone employable five years later.

Where the real tension sits: GCCs and the "outcome-based" shift

One trend worth watching closely is what's happening inside India's Global Capability Centres. These aren't the offshore back-offices of a decade ago — GCCs leased a record 9 million square feet of office space in early 2026 alone, and they're increasingly running AI-native operations rather than legacy BPO functions. That's genuinely new demand, not a replacement for old demand.

At the same time, IT services providers are restructuring how they bill clients — moving away from the old "bodies on a project" (FTE-based) model toward outcome-based, risk-sharing contracts. That single shift, more than any AI headline, may be the biggest structural change underway. When you stop paying for hours and start paying for results, headcount stops being the natural way to scale a services business. You can grow revenue significantly without growing people at anywhere near the same rate — which is more or less exactly what the FY26 numbers show happening already.

The skills gap is the actual bottleneck, not the technology

Here's a statistic worth sitting with: despite AI-related job demand nearing a million roles, only around 16% of India's IT professionals are currently considered AI-skilled, according to figures cited from the Ministry of Electronics & IT. Separately, industry estimates suggest roughly half of AI/ML roles remain unfilled. That's not a story about AI destroying jobs. That's a story about a mismatch — plenty of demand, not enough supply, and a training pipeline that hasn't caught up.

This is actually the more hopeful part of the picture, because mismatches are fixable in a way that structural unemployment isn't. The 2026 India Skills Report found that over 90% of Indian employees have already started using generative AI tools at work, and national employability has climbed from about 46% in 2022 to over 56% now. People are adapting faster than the doom narrative gives them credit for. The gig and freelance workforce tied to tech skills is projected to keep growing too, which suggests the labor market is finding new shapes for work rather than simply shedding it.

So — more jobs, or fewer?

Honestly, the net-number framing might be the wrong question. The more useful one is: which jobs, held by whom, doing what kind of work? Under that lens, here's roughly where things stand:

• Total headcount in India's tech services sector is very likely to keep growing, just slowly, and increasingly decoupled from revenue growth.
• Genuinely new categories of work — AI engineering, prompt and agent design, AI governance, GCC-based AI operations — are real and growing fast, but concentrated among people who already have strong technical foundations.
• Entry-level and routine cognitive work is shrinking as a share of total jobs, which threatens the traditional path India's services industry used to offer as a ladder into the middle class.
• The binding constraint right now isn't AI's capability, it's how fast people can be trained to work alongside it — and that's a policy and education problem as much as a technology one.

If India gets the reskilling piece right — and NASSCOM's own "Roadmap for Job Creation in the AI Economy" frames this almost exactly as a fork in the road, with a possible swing of nearly 4 million jobs (a 1.5 million loss on one path, a 2.5 million gain on the other) — the sector comes out ahead. If it doesn't, the same technology produces a much less comfortable outcome: fewer entry points, a narrower band of winners, and a services industry that grows on paper while employing fewer people to get there.

The technology isn't deciding this. The next few years of policy, corporate training budgets, and how seriously India takes the "who trains the freshers now" question — that's what will decide it.