AI will cut 60% of jobs.

Our finding

Why we rated this “False”

What the 60% actually refers to

This figure has a clear origin, and its meaning is stated plainly in the research it comes from.

IMF analysis estimates that roughly 60% of jobs in advanced economies are exposed to artificial intelligence. [1]

Exposure is a measure of contact, not of loss. A job is exposed when AI can perform some part of what it involves. That can play out in three quite different ways: automation, where a task is taken over entirely; augmentation, where the technology makes a worker faster or more capable; or productivity gain, where output rises without the role changing much.

Only the first of those subtracts a job. The 60% figure counts all three together.

The advanced-economy qualifier

There is a second problem with applying this number to India.

The estimate covers advanced economies, whose employment is weighted heavily towards office-based, information-processing work — precisely the roles with the most AI contact. India's employment structure is different, with a much larger share in agriculture, construction, retail and other work where exposure is considerably lower.

The IMF's own analysis puts exposure in emerging markets and low-income countries well below the advanced-economy figure. Quoting the advanced-economy number in an Indian context transfers an estimate to a labour market it was not measuring.

The net projection

If the question is what happens to total employment, the World Economic Forum has published a projection that addresses it directly.

It expects 170 million jobs created and 92 million displaced globally by 2030 — a net gain of about 78 million. [2]

That is a forecast, and forecasts about technology and employment have a poor historical record in both directions. But it is the projection being drawn on, and it points to net creation rather than net loss.

Where the claim goes wrong

"Exposed to AI" becomes "cut by AI." A measure of how many jobs the technology will touch is presented as how many will be eliminated.

The advanced-economy scope is also dropped, applying a figure to India that was calculated for a different set of labour markets.

What the research does warn about

The IMF analysis is not reassuring, and it would be wrong to present it as though it were.

Its central concern is distribution rather than totals. AI is expected to widen inequality — between workers who can use it and those displaced by it, and between countries positioned to deploy it and those that are not. Some workers will lose jobs and will not find comparable replacements, even if aggregate employment grows.

There is also a specific risk to entry-level work. Routine tasks that have traditionally given graduates their first job are among the most automatable, which is a direct concern for the audience this claim was addressed to.

So the honest reading is that AI poses serious and unevenly distributed risks to employment, without supporting the specific figure of 60% job losses.

In short

60% is the IMF's estimate of jobs *exposed* to AI in advanced economies — a category that includes augmentation and productivity gains, not just automation. The WEF projects a net global gain of about 78 million jobs by 2030. The real risks are distributional, and they are not what this number measures.