87% of Americans want a human to approve AI layoffs

A June survey of 500 US adults found 87% want a human to sign off on any AI-recommended layoff, and most doubt AI is the real reason at all.

87% of Americans want a human to approve AI layoffs

Companies spent 2026 discovering that “AI” is a wonderful word to say in a layoff announcement. It signals technological seriousness, reassures investors that the cuts are strategy rather than distress, and conveniently makes a painful decision sound like progress. A new survey suggests the public has stopped buying it.

A national poll of 500 US adults, run in June 2026 by Kolmogorov Law through the Pollfish platform and written up by VMblog on June 3, found near-unanimity on one point: 87% say a human manager should be required to review any layoff recommended by AI before it takes effect. Only about 8% disagree. In survey work, where consensus is rare and “no opinion” usually wins, that is roughly as close to a mandate as public sentiment produces.

The public has learned the word “AI-washing”

The more revealing numbers are the ones about belief. Asked how often AI is the real reason behind layoffs that blame AI, only 15% said “almost always.” A 45% plurality said it was the true driver “about half the time,” and another 40% said AI is rarely the real cause and usually just an excuse for cost-cutting a company had already decided on. Put differently, 85% of Americans think that, at best, AI is only part of the story.

That skepticism gets sharper when it’s personal. Asked whether they would believe their own employer if it announced cuts and pinned them on AI, 41% said no — outnumbering the 36% who said yes. Nearly two in three workers, in other words, would not take the explanation at face value. This is not paranoia; it’s pattern recognition. OpenAI’s Sam Altman has said that “almost every company that does layoffs is blaming AI, whether or not it really is about AI.” Oxford Economics concluded earlier this year that firms “don’t appear to be replacing workers with AI on a significant scale.” Deutsche Bank analysts even coined a phrase for the year ahead: “AI redundancy washing.”

The canonical cautionary tale remains Klarna, which boasted that an AI assistant was doing the work of 700 customer-service agents — then quietly began rehiring humans, with its CEO conceding the company “went too far” and that cost-cutting had bruised quality. The arc is now familiar: the AI-replacement claim arrives loud and early, the correction arrives quiet and late.

What people want, and where the law is going

Beyond the headline 87%, the survey found 76% want companies legally required to disclose when layoffs are driven by automated systems, and 62% say the company that used the AI — not the vendor that built it (13%), and certainly not “no one, it’s just software” (12%) — should be accountable when an AI-driven layoff turns out to be unfair or unlawful. These are supermajorities, not partisan splits.

California is the first state seriously testing whether to turn that sentiment into statute. The original “No Robo Bosses Act” (SB 7) passed both houses in 2025 before Governor Newsom vetoed it as overly broad; its author reintroduced a narrower Senate Bill 947 in 2026, barring employers from relying solely on AI to fire or discipline workers and requiring human review of those decisions. All of it layers on top of California’s existing Cal-WARN Act, which already demands 60 days’ notice for covered mass layoffs — an obligation that does not evaporate because an algorithm recommended the cut.

The gap that makes the dodge work

Here is the uncomfortable part. For all their skepticism, most workers are unequipped to act on it. Only 23% said they were “very confident” they understood their rights to notice and severance; 37% were not confident at all. At the same time, 46% are worried AI could eliminate their own job within three years. High concern, low knowledge — that’s the exact combination that makes an unverifiable “AI” explanation so effective. A worker who doesn’t know what notice they were owed is unlikely to ask whether the law was followed.

The full press release frames the verdict cleanly: the public wants a human in the loop and a paper trail behind the decision. The skepticism is already here. What’s missing is the leverage — and, for now, the homework. If you take one thing from this survey, let it be the boring one: find out what your jurisdiction owes you before the meeting invite with no agenda lands in your calendar.

Sources

Keep reading