AI Roles Are Up 25% While IT Hiring Falls: What the Indian Data Actually Says

Most commentary about AI and employment is speculation about a future nobody can see. The Naukri JobSpeak index is not that. It counts job listings posted on India's largest hiring platform and compares them with the same month a year earlier — a direct measurement of what employers are actually advertising for.

Here is what the June 2026 data shows, and what it reasonably implies for someone deciding where to point their next two years.

The numbers

The overall index stood at 3027 in June 2026, against 2854 in June 2025 — an increase of 6%. Beneath that headline, the sectors diverge sharply.

And by function, the growth is concentrated somewhere many readers will not expect:

Reading the IT paradox properly

“IT down 3%, IT and information security up 18%” looks contradictory. It is not. The first figure counts the sector — every role at IT services companies, including the enormous base of routine delivery work. The second counts the function — technology and security roles wherever they sit, including inside banks, hospitals, manufacturers and media companies.

Put together, the two lines tell a coherent story: technology work is not shrinking. It is relocating. It is moving out of the large services firms that historically aggregated it, and into the companies that use it.

The experience curve is inverted

Conventional wisdom holds that a slow market punishes senior people first, because they cost more. The June data says the opposite.

The weakest band by a clear margin is 4 to 7 years, at just 2%. That is precisely the band whose work is most exposed: experienced enough to be expensive, not yet senior enough to be making decisions, and doing tasks that sit squarely in the range current tooling handles well.

If you are in that band, this is the number to take seriously. Not as a reason for alarm — hiring is still growing — but as a reason to be deliberate about what you are becoming good at over the next two years.

Where the growth actually is geographically

Bengaluru grew 7%. Kolkata grew 12%, Hyderabad 11%, Chennai 10%. And the smaller cities outpaced all of them: Bhubaneswar +21%, Indore +15%, Coimbatore +10%.

This is a structural shift, not a blip. Distributed work made second-tier cities viable for employers, and the cost differential makes them attractive. For anyone weighing a move — or weighing whether to move back — the assumption that opportunity requires Bengaluru is weaker than it was three years ago.

What “AI/ML roles” actually means, and what it does not

A 25% growth figure invites a bad conclusion: that everyone should retrain as a machine learning engineer. That is the wrong lesson, for two reasons.

First, the base is small. Twenty-five per cent growth on a small number of specialist postings is a much smaller absolute quantity of jobs than 6% growth across the whole white-collar market. Chasing the percentage without looking at the base is how people end up competing for a few thousand roles against candidates with doctorates.

Second, the growth is not confined to model building. The category includes data engineering, MLOps, annotation and evaluation, applied product roles, and the governance and safety work that regulated industries now require. Much of it is closer to careful engineering and domain judgement than to research.

What to do if you are in the 4–7 year band

Concrete, in rough order of return on effort.

1. Get fluent with the tools in your own function, not in the abstract

A finance analyst who can move a three-day reconciliation to an afternoon is worth more than one who has completed a generic AI course. The differentiator is not knowing what a transformer is. It is having demonstrably compressed a real process at a real employer.

2. Move toward judgement, away from production

The tasks under most pressure are the ones with a clear right answer and a defined procedure. The tasks that hold value are the ones requiring a decision about ambiguity, responsibility for an outcome, or a relationship with a person. Steer your next role toward the second kind.

3. Look outside your sector before you look outside your function

Given technology functions growing 18% while the IT sector falls 3%, the same skills carry a better price inside insurance, healthcare or manufacturing than inside a services firm. That is a smaller and less risky jump than retraining.

4. Write down what you have actually done, with numbers

None of this reaches anyone if your CV says “responsible for reporting”. Employers hiring for the growing categories screen on demonstrated outcomes. Most people have the outcomes and have simply never recorded them. If your applications are getting no response at all, the cause may be mechanical rather than substantive — our guide to why parsing software rejects CVs covers the ten-minute test that tells you which.

5. Do not quit to study full time

The market is growing at 6%, not contracting. Leaving a paid role for a certificate is an expensive way to acquire something you could acquire on the job, and it removes the one thing employers in this data are actually buying: evidence you have done it in a real workplace.

The honest caveats

One platform's listings are not the entire labour market. Naukri is the largest in India but skews toward white-collar, urban and formally advertised roles; senior hiring often happens through networks and never appears as a listing. A single month can move on one large employer's decisions, which is why the quarterly figures matter more than any one month.

Job postings also measure demand advertised, not positions filled. A 25% rise in AI/ML listings is partly employers competing for a genuinely thin supply of qualified people — the same role reposted is counted again.

None of that overturns the direction. It argues for reading the divergence between sectors and functions as the signal, rather than any single percentage.

The bottom line

The Indian white-collar market in mid-2026 is growing modestly overall, contracting in traditional IT services and banking, and growing sharply in technology work embedded inside other industries. Freshers and senior people are doing better than the middle. Smaller cities are outpacing the metros.

For the individual, the actionable version is short. Stay employed. Get genuinely good at the current tools inside the domain you already understand. Aim at work requiring judgement rather than production. Consider a sector move before a career change. And write down your results with numbers attached, because none of it counts until someone can see it.

Frequently asked questions