AI Gets the Blame for the Layoffs. Does It Deserve It?

Every few weeks another company announces job cuts and points at artificial intelligence. It has become the explanation of record for a shrinking payroll, and for an anxious workforce it confirms the oldest fear about this technology, that the machine is here for the desk. The claim is everywhere. The evidence underneath it is thinner than the headlines suggest.

Start with what is true. The outplacement firm Challenger, Gray and Christmas counted more than 101,000 US job cuts attributed to AI through June of this year, and for four straight months AI was the single most cited reason employers gave. The share of layoff announcements that named AI climbed from about 7 percent in January to roughly 40 percent by May. As a measure of what companies are saying, the trend is real and steep.

As a measure of what is actually happening, it is doing a lot of quiet work. "A company saying AI is why they're doing layoffs doesn't necessarily mean that's actually why," Daniel Zhao, chief economist at Glassdoor, told CNBC, adding that he is skeptical the current cuts reflect genuine efficiency gains. The reason for his doubt is simple. If AI were truly replacing workers through productivity, you would expect to see productivity rise. So far it mostly has not. Goldman Sachs reported earlier this year that it still found no meaningful relationship between AI adoption and productivity across the economy as a whole.

Why "AI did it" is such a convenient line

There is a name for the gap between the saying and the doing. Analysts have taken to calling it AI washing, and the pattern is consistent. A company cites AI as the engine of a smaller headcount, then insists in the same breath that it is not really replacing people. Even Sam Altman, whose company sells the tools in question, has observed that nearly every firm conducting layoffs now reaches for AI as the reason, whether or not it is the cause.

Blaming the software is useful in ways that have nothing to do with software. It reframes a cut made because a company over-hired, or because borrowing got expensive, or because demand softened, as a forward-looking bet on technology rather than a retreat. It flatters the share price. It spares executives the less flattering admission that they misjudged how many people they needed.

None of this means AI is harmless to employment. Economists at the National Bureau of Economic Research have found that finance chiefs expect to make far more AI-driven cuts over the coming year, so the pressure is real and building, even where the promised productivity has yet to show up. Particular roles, especially routine and entry-level ones, are genuinely exposed. The honest position is that we are early, the data is noisy, and the confident attributions are running well ahead of what anyone can actually measure.

So treat the next "AI took these jobs" announcement as a claim, not a finding. Sometimes it will be true. Often it will be a tidy story laid over an untidier decision. If you want to know whether the machines are coming for the work, watch productivity, not press releases, and notice who benefits from the explanation. We have looked at the same fault line in how hiring itself is being handed to automated systems and in the broader worry that the tools are changing us even as we use them.