In twenty-four years in hiring, I have watched plenty of technologies take the blame for cuts they did not make. So when every headline this year says AI is killing India's entry-level jobs, I do not argue with it. I ask the more uncomfortable question. Is AI actually doing the cutting, or is an ordinary market correction wearing AI as a costume?
Mostly the costume. And the costume is doing real damage, because it tells a whole generation the door is shut when the door has only moved.
The layoffs are real. The reason on the press release usually is not.
Start with the cuts, because this is not a comfort piece. They are real, and they are large. Read the honest number, net headcount and not gross intake, and the three biggest Indian IT firms cut about 64,000 roles in FY24, and TCS alone shed around 25,000 across 2025. Nobody who lost a seat cares whether a spreadsheet or an algorithm took it.
But the reason matters for what you do next, and the reason does not hold up. The heavy cutting started in 2023 and 2024, when AI tools were barely usable inside a real enterprise. The actual driver was older and duller. Through the pandemic digital boom, India's IT firms hired at record pace to chase a surge of client work. Then global interest rates rose, their clients in the US and Europe froze discretionary spending, revenue growth stalled, and the bench filled up with people there was no longer billable work for. So the firms did what any business does in a downturn: they stopped hiring and trimmed the surplus they had built. None of that is artificial intelligence. All of it is the ordinary business cycle.
So why does every press release say AI?
"We are going AI-first" sells better than "we over-hired"
Because the honest version is embarrassing and the AI version is visionary. "We misjudged the cycle and hired too many people" makes a chief executive look careless. "We are restructuring into an AI-first company" makes the same person look like a prophet. Same cuts, better story, higher share price. India's own IT leaders are among the keenest to tell it, because an AI-first transformation reassures investors far better than admitting their clients simply stopped spending.
This is not a fringe theory, and even abroad the people selling the AI revolution are saying it out loud. Nvidia's Jensen Huang has asked executives to stop telling workers they were fired because of AI. When the single biggest salesman of the AI boom is begging people to stop blaming AI for their layoffs, you should listen.
Same cut, better costume. "We over-hired" makes a CEO look careless. "We are going AI-first" makes the same CEO look like a prophet.
Part of it is AI, though, and it lands on exactly one rung
Here is where I refuse the easy answer, because dismissing AI completely is as wrong as blaming it for everything. There is a real AI shift under the noise, and it is hitting one place hardest: the bottom rung.
Routine first-draft code, basic data work, first-line support, the simple repetitive tasks we used to hand a fresher in their first six months, are exactly what today's tools do well enough. In India, entry-level IT roles have already shrunk by 20 to 25 percent as automation takes that work over. That part will not bounce back when rates fall, because it is the work changing shape, not the budget.
The correction explains how many people lost their jobs. AI explains why the first job is the one that disappeared.
India is the sharpest case study in the world
Nowhere is the double story clearer than here, because you can watch firing and hiring happen in the same week. Net headcount is down, and yet the big firms still plan to onboard tens of thousands of freshers, around 80,000 between them, in the same year. That is not an industry being wiped out by machines. It is an industry swapping one kind of worker for another, letting generalists go while it fights over a narrower, better-prepared slice. AI, machine learning and cloud now sit in the job description where a generic engineering degree used to be enough.
And here is the detail that breaks the whole "AI took the jobs" story. While we lay off thousands, we cannot find the people we actually need. AI talent demand is heading past a million, only about 16 percent of the current IT workforce is genuinely AI-skilled, and the gap between what employers want and what they can find is around half. We are running a jobs scare and a talent shortage at the same time. That is not what a machine eating jobs looks like. That is a skills mismatch, and a skills mismatch is an education problem first and an economic one second.
The number nobody puts in the headline: employability is going up
If the story were simply "AI eats Indian graduates," employability would be falling off a cliff. It is climbing. The India Skills Report 2026 puts graduate employability at 56.35 percent, up from about 55 percent a year earlier and from the mid-forties four years ago, with computer-science and IT graduates at the top and women overtaking men for the first time in five years. Global Capability Centres are hiring hard across Bengaluru, Hyderabad and Pune.
The opportunity is real. It is just flowing to a narrower, better-prepared slice than before. The graduate who can only do what the tool already does is the most exposed person in the market. The graduate who can direct the tool, inside a field they actually understand, has never been more wanted.
The most exposed person in this market is the one who can only do what the tool already does.
Who the costume actually protects
Ask why the costume fits so comfortably, and the answer is uncomfortable. Blaming AI lets almost everyone off the hook.
It lets the company look like a visionary instead of a careless over-hirer. It lets the college avoid admitting its syllabus is a decade out of date. It lets the policymaker point at a global force instead of a local failure. And it hands the student a reason to give up instead of a reason to retrain. Four parties walk away relieved, and only one is left paying the bill: the graduate who believed the door was shut and stopped knocking. That is why the costume is so popular, and exactly why it is worth tearing off.
What to do once you stop believing the costume
If you are a graduate, the worst move is to panic, because fear turns into paralysis and paralysis is fatal in a market that is re-sorting. Treat AI fluency as the floor, not the ceiling, then pair it with a real domain so you are competing in a field instead of competing with a chatbot. Build a body of owned work to stand in for the first job that used to teach you on the job. Look well beyond legacy IT services. And accept that retraining is the career now, not a break from it.
For institutions, the work is sharper, and I say this having spent years inside engineering colleges across India.
- Pull AI literacy out of the elective corner. Employers screen for it, so it belongs running through every subject like reading and arithmetic, not bolted on as one course at the end.
- Rebuild the first job that AI dissolved. The entry rung where a graduate used to become a professional is vanishing, so part of that apprenticeship has to happen before graduation: real client briefs, judgement-heavy internships, and assessment that rewards problem framing, communication and ownership over output a model can already produce.
- Sell direction, not alarm. The most useful thing a placement cell can do this year is stop forwarding the doom and start teaching agency.
The honest bottom line
Is AI replacing entry-level jobs in India? Partly, and permanently, at the bottom rung. That part is real and deserves a serious answer. But most of what gets blamed on AI is a market correction in costume, and mistaking the costume for the truth points graduates and colleges at exactly the wrong conclusion. The door has not closed. It has moved, and it has narrowed, and the people who learn where it went will walk straight through it.
Reading the signal, not the costume
This is the work we think about every day at Skillencio: telling whether a graduate is genuinely ready to deploy, not just holding a degree, because that is the line the new market actually hires on. You do not need us to start. You need to read the real labour-market signal instead of the headline, and build for the market that is arriving rather than the one that left.
Common questions
Will AI replace entry-level jobs in India?
Partly. AI is taking over the routine entry-level work, and India's entry-level IT roles are already down 20 to 25 percent. But most of the headline layoffs are an ordinary market correction, record over-hiring unwound when client budgets and interest rates turned, not AI. The correction explains how many jobs went; AI explains why the first rung is the one that vanished.
If AI is taking jobs, why is graduate employability in India rising?
Because the work is being re-sorted, not erased. Graduate employability reached 56.35 percent in the India Skills Report 2026, and GCCs and AI-native firms are hiring hard. The jobs are flowing to a narrower, better-prepared slice: the graduates who can direct AI inside a real domain, not just do what the tool already does.
What should students learn to stay employable as AI reshapes hiring?
Treat AI fluency as the floor, then pair it with a real domain so you compete in a field instead of against a chatbot. Build a portfolio of owned work to stand in for the vanishing first job, and look well beyond legacy IT services to GCCs, product firms and AI-native startups.
