TL;DR: Q1 2026 tech layoffs hit 78,557, with 47.9% attributed to AI in corporate announcements. A 6,000-executive NBER survey found 90% reported no measurable AI impact on employment over three years. A pending federal bill would force quarterly AI-layoff reporting to the Department of Labor.

Key Takeaways

• Nearly 80,000 tech workers were laid off in Q1 2026, with 47.9% of cuts publicly attributed to AI and workflow automation.
• An NBER working paper surveyed roughly 6,000 senior executives and found 90% reported no AI impact on employment over the past three years.
• The AI-Related Job Impacts Clarity Act (S.3108) would require covered employers to file quarterly AI-layoff reports to the U.S. Department of Labor.
• Union membership hit a historic low of 9.9% in 2024, and union members are the workers most likely to file for unemployment benefits after a layoff.
• OpenAI CEO Sam Altman publicly acknowledged the "AI washing" dynamic at the India AI Impact Summit.

AI Washing and the 90% Layoff Reporting Gap

Tech companies attributed nearly half of their Q1 2026 layoffs to AI. A Federal Reserve-backed survey of almost 6,000 executives says 90% of them saw no workforce impact from AI at all. Both cannot be true.

The phrase "AI replaced my job" now shapes severance negotiations, unemployment eligibility interviews, and retraining pathways. It also shapes how investors price firms and how policymakers write rules. But no regulator today has the authority to verify whether a single layoff was actually caused by AI adoption. That gap is what this piece is about.

The Numbers That Don't Match

Between January 1 and April 2026, 78,557 U.S. tech workers were laid off, according to industry aggregators. Of those cuts, 37,638 (47.9%) were attributed to reduced need for human workers because of AI and workflow automation. The attribution came from company statements and analyst trackers counting those statements.

In the same quarter, the National Bureau of Economic Research published working paper w34984, Artificial Intelligence, Productivity, and the Workforce: Evidence from Corporate Executives. The paper surveys roughly 6,000 senior executives across U.S., U.K., German, and Australian firms. More than 90% of managers reported no AI impact on employment over the past three years. 89% reported no change in productivity. Looking forward, the same executives predicted only a 0.7% AI-driven cut to employment over the next three years.

The gap between "47.9% of tech layoffs were AI" and "90% of executives saw no AI workforce impact" is not a rounding error. It is two incompatible accounts of the same quarter, one from press releases and one from a confidential executive survey. Which one describes reality matters for every downstream policy response.

Who Benefits When a Layoff Gets Branded as AI?

Framing a layoff as AI-driven changes three things at once. It repositions the company as modern rather than declining. It reduces public sympathy for displaced workers, because "technology moved on" reads as inevitable rather than managerial choice. And it shifts scrutiny away from weaker fundamentals, like failed product bets, overhiring during 2021 and 2022, or end-of-cycle cost cuts.

OpenAI CEO Sam Altman said at the India AI Impact Summit: "I don't know what the exact percentage is, but there's some AI washing where people are blaming AI for layoffs that they would otherwise do, and then there's some real displacement by AI of different kinds of jobs." The admission from the CEO of the most valuable AI company in the world is worth reading twice. He did not deny displacement. He said some fraction of the attribution is narrative.

So the question worth asking in 2026 is not "is AI replacing workers?" It is "which layoffs does AI actually explain, and which layoffs just needed a more defensible headline?"

What the Clarity Act Would Force Into Public View

On November 19, 2025, Senators Josh Hawley (R-Mo.) and Mark Warner (D-Va.) introduced S.3108, the AI-Related Job Impacts Clarity Act. The bill would require covered employers to file quarterly reports with the U.S. Department of Labor. The reporting categories are specific.

The bill sits in the Senate Committee on Health, Education, Labor, and Pensions. It has bipartisan authorship but no scheduled markup as of this writing. Its passage is uncertain. What matters for this discussion is the category of data it would create: a comparable, quarterly, employer-filed record of AI employment impact, held by a federal labor regulator.

Today, no such record exists. Every claim about AI-driven layoffs is either self-reported by the firm in a press release or estimated by a third party counting those press releases. The NBER paper is the closest thing to an independent baseline, and it contradicts the press-release aggregation directly.

Why Union Decline Makes the Reporting Gap Worse

Nearly 75% of Americans who lose their jobs do not apply for unemployment insurance benefits. One of the strongest predictors of whether a laid-off worker files is union membership, because unions walk members through the process. Union density in the U.S. reached a historic low of 9.9% in 2024, according to Bureau of Labor Statistics data.

The labor-rights consequence is concrete. When a layoff is publicly framed as AI-driven, a displaced worker interviewing for unemployment does not have a union representative cross-checking the framing against what the employee actually did. If the official narrative says "your job was automated," the administrative record often ends there. This is not theoretical. States with the highest proportion of unclaimed benefits are also states with the lowest union density.

So the reporting gap is not just an academic dispute between NBER and corporate press teams. It is a structural asymmetry in who gets to describe a job loss, and what that description does to the worker's next six months.

The Future-Projection Trap

The same NBER survey that found 90% of executives reporting no past AI impact also found those executives expect roughly 0.7% AI-driven employment cuts over the next three years. That is a small number. It is also a planning assumption that would not, on its own, justify the scale of Q1 2026 cuts.

The Harvard Business Review put it plainly in January 2026: companies are laying off workers because of AI's potential, not its performance. This is the trap. "We had to cut now because AI will replace the function soon" is a defensible management narrative, but it is non-falsifiable in the short term. Investors accept it. Boards accept it. Labor regulators have no standing to question it. And when the three-year window closes and the projected AI displacement is smaller than the actual layoff count, no one reopens the original attribution.

When the AI Layoff Narrative Doesn't Apply

Not every 2026 layoff is AI washing. Some job functions are genuinely being absorbed. Pattern recognition for customer support routing, first-pass legal document review, low-complexity image editing, formulaic copy production: these categories have real automation exposure. A worker in one of these roles reading this piece should not walk away thinking their displacement was fake.

But the categories where AI actually replaces labor are narrower than the current coverage suggests, and the concentration is uneven. The cases where "AI replaced this role" is a defensible claim share three markers.

• ☐ The replaced function was already highly routinized before 2024.
• ☐ The firm has deployed a specific, named AI system and can point to measurable output from it.
• ☐ The headcount reduction lines up with the specific team where that system was deployed, not with an across-the-board restructuring.

If any of those three are missing, the "AI did it" framing is doing narrative work, not descriptive work. That distinction is what the Clarity Act would force into public disclosure.

FAQ

Is the NBER survey representative of U.S. tech companies?

The NBER paper (w34984) covers roughly 6,000 senior executives across U.S., U.K., German, and Australian firms, weighted toward mid-and-large enterprises. It is the largest executive-level AI impact survey published in 2026. It is not tech-sector-exclusive, which is both a caveat and a reason its finding matters: even including sectors most likely to claim AI impact, 90% saw no workforce effect.

Has the AI-Related Job Impacts Clarity Act passed?

No. As of April 2026, the bill (S.3108) sits in the Senate Committee on Health, Education, Labor, and Pensions. No markup has been scheduled. Bipartisan authorship improves its odds but does not guarantee passage.

What's the difference between AI washing and real AI displacement?

Real displacement can be traced to a specific deployed system, a measurable change in output, and a headcount change concentrated in the team using that system. AI washing applies the AI label to layoffs that share none of those markers and would have happened regardless.

Why does this matter for workers already laid off?

The framing of a layoff affects unemployment filing rates, retraining eligibility, and severance negotiation leverage. Workers laid off under an "AI automation" justification face a harder time contesting the decision than those laid off for documented performance or redundancy reasons, even when the underlying facts are identical.

Conclusion: Next Steps

The honest summary of 2026 so far is that the AI layoff narrative is running well ahead of the workforce data. Aggregated tech layoff trackers report 47.9% of cuts as AI-driven. An NBER executive survey covering 6,000 firms says 90% saw no workforce impact. Federal reporting does not yet exist to reconcile the two. Workers pay the cost of the gap.

If you are a policy analyst, labor researcher, or journalist covering this beat, the concrete action this quarter is to read the NBER working paper directly before citing any layoff aggregator that attributes cuts to AI. If you are an HR leader or workforce strategist inside a company currently considering AI-attribution framing for a restructuring, the question worth putting to your general counsel is what the Clarity Act's quarterly reporting regime would require you to disclose if it passed in the next session. Assume it will pass, and check whether your attribution survives the audit trail.