The Federal Government's Official AI Layoff Count Is Zero
The Algorithm has, by most private tallies, cost more than 100,000 Americans their jobs this year. The federal government's own books show none of it. Unemployment insurance data doesn't ask why someone was let go. The WARN Act, which requires big employers to give advance notice of mass layoffs, routes that information to state and local governments and stops there — it never reaches Washington. Independent contractors, remote workers, and staff at smaller companies fall outside the count entirely. As far as the official record is concerned, the layoffs attributed to AI simply are not happening.
Three bills now sitting in Congress exist to fix a gap that, one would think, might have been worth noticing before the layoffs started.
Three Different Answers to "How Would We Even Know"
Senators Mark Warner and Ted Budd introduced the bipartisan Workforce Transparency Act in April, which would let the Department of Labor collect voluntary data on where and how companies use AI, so lawmakers might eventually craft policy based on evidence rather than vibes. "We're already seeing it have a measurable effect on the U.S. workforce," Warner noted, correctly, while proposing that companies be asked nicely to mention it. Representatives Steven Horsford, Sara Jacobs, and James Moylan went further with the AI-Related Job Impacts Clarity Act, which would require, not request, that major employers and federal agencies disclose AI-attributed layoffs to the Labor Department. "Behind closed doors, with no disclosure and no accountability, entire livelihoods are being erased," Horsford said, describing a phenomenon his own bill concedes nobody can currently measure.
Representative Greg Casar's AI Tax and Work Protection Act skips the counting problem and goes straight to consequences: a tax on AI companies' token sales and product revenue that automatically rises once national unemployment climbs past five percent, funding a new Work Protection Administration to employ whoever the automation displaces.
Meanwhile, the Private Sector Kept Its Own Spreadsheet
The Machine did not wait for Congress to catch up. Layoffs.fyi and other trackers put 2026's AI-cited job cuts well past 100,000, with the share of layoff announcements naming AI or automation as a factor climbing from roughly 7 percent in January to 40 percent by May — a rate of increase that outpaces the technology's own actual capability gains, which is a polite way of saying it has become a rationale companies deploy because investors reward it. Challenger, Gray & Christmas confirmed May as the sector's worst single month for layoffs in years, AI cited most often as the reason.
Sources within the Labor Policy Community confirm that a spreadsheet run by a jobs website has, for the moment, a clearer picture of AI's effect on employment than the United States government does.
Filed Under: We'll Get Back to You on That
None of the three bills has passed. Even the most ambitious of them, Horsford's mandatory-disclosure act, would only begin producing federal numbers after enactment — meaning the government's authoritative count of AI-driven layoffs, whenever it finally arrives, will describe a labor market that has already moved on twice. In a development that will surprise no one who has been paying attention, the private trackers will keep counting in the meantime, because someone has to, and it certainly isn't going to be the WARN Act.
The gateway, it appears, still requires a form. Congress just hasn't agreed on which one.
Sources: Equitable Growth · Sen. Warner · Rep. Horsford · TechCrunch
