Are Mass AI Layoffs Actually Happening? The Data vs. The Expectations

What the Hiring Data Actually Shows

Despite the stark executive sentiment, the most comprehensive empirical studies of U.S. hiring tell a more measured, compositional story.

The Harvard Business School Evidence

A landmark working paper from Harvard Business School, "Displacement or Complementarity? The Labor Market Impact of Generative AI," analyzed nearly all U.S. job vacancies from 2019 through March 2025 . The study, led by Professor Suraj Srinivasan, found a clear shift in what employers want, but not an overall collapse in demand.

After the public launch of ChatGPT in November 2022, job postings for occupations dominated by structured and repetitive tasks—the kind most exposed to automation—declined by 13% . Simultaneously, demand for roles requiring analytical, technical, or creative work—positions where AI can augment human capability—grew by 20% .

The researchers documented a clearly heterogeneous effect: generative AI is reducing demand and skill requirements in structured cognitive-task jobs while increasing both in positions that involve human-AI collaboration . As the authors put it, the evidence points toward compositional shifting rather than wholesale job elimination in the available data .

The Federal Reserve Perspective

Federal Reserve research amplifies this cautious picture. A Federal Reserve Board analysis of firm-level data found no evidence of a reduction in job postings for industries or firms with higher levels of AI adoption . The study concluded that the overall national slowdown in job postings after the pandemic recovery does not appear to be driven, even modestly, by AI .

The Dallas Fed's January 2026 Beige Book noted that most companies using AI reported it had not affected their employment levels, though about a quarter did expect AI to decrease their need for workers over the next few years . Research from the Richmond Fed's CFO Survey similarly found little evidence that firms have experienced or anticipate near-term AI-driven employment declines, even as companies—especially larger ones—anticipate reshuffling their workforce away from routine clerical jobs toward more skilled technical roles .

A Reality Check on the Numbers

The gap between CEO expectations and observed reality is stark. While 99% of CEOs say they are ready for AI layoffs, a National Bureau of Economic Research working paper surveying 750 U.S. CFOs found that less than half (44%) actually plan on some AI-related job cuts this year. When scaled to the broader economy, that amounts to roughly 0.4% of all roles, or about 502,000 positions out of approximately 125 million—a notable increase from the estimated 55,000 AI-attributed layoffs in 2025, but still a rounding error against the total workforce .

The Richmond Fed's CFO survey showed that 59% of firms actually planned to increase their workforce in 2026, with layoff plans driven more by demand uncertainty than by AI specifically . This suggests that while executive sentiment is overwhelmingly bearish on AI's employment impact, firms' actual hiring and firing plans remain dominated by broader economic conditions.

The Human Toll: Employee Well-Being and the Entry-Level Squeeze

Even if the macroeconomic data doesn't show a jobs apocalypse, the fear of one is already causing measurable distress.

Psychological Impact on Workers

Peer-reviewed research published in 2025 found that AI-related job anxiety significantly and negatively predicts life satisfaction, with negative emotions fully mediating that relationship . A separate study on AI-driven job displacement among Indian IT professionals documented emotional shock, erosion of professional identity, chronic anxiety, and perceived organizational betrayal among those who lost roles to automation .

Researchers at the University of Florida have gone further, proposing a new clinical construct called AI Replacement Dysfunction (AIRD) to describe the stress linked to sustained fears of AI-related job loss . Symptoms can include anxiety, insomnia, paranoia, denial of AI's relevance, loss of identity, feelings of worthlessness, resentment, and hopelessness . AIRD is not a formal DSM diagnosis, but it is increasingly treated as a legitimate clinical concern in 2026 mental-health discourse .

Employee surveys paint a troubling picture: Modern Health's 2026 Workforce Mental Health Report found that nearly 7 in 10 US employees (69%) now believe AI will lead to layoffs at their own company within three years, and almost half (49%) are personally afraid of losing their jobs to AI tools and automation . An ADP Research survey of more than 39,000 workers across 36 countries found only 22% of employees worldwide strongly agreed their job was safe from elimination .

Entry-Level Workers and the Shrinking On-Ramp

The most immediate labor-market burden appears to be falling on workers in entry-level structured-task roles. The Harvard Business School study's core finding—the 13% decline in postings for routine, automatable occupations—directly affects positions that have historically served as on-ramps to professional careers .

In occupations with high AI exposure, entry-level hiring has declined specifically as large language models have proliferated at scale . Rather than displacing workers uniformly, the research suggests AI is creating a two-tier workforce in which experienced professionals see rising productivity and expanded roles while entry-level workers face shrinking opportunities and stagnant prospects . The result is a bifurcation of the labor market, with AI automating execution-based tasks while amplifying the value of experienced judgment—and narrowing the traditional paths for new entrants to build that experience .

What This Means Moving Forward

The data in 2026 presents a paradox of its own. The near-unanimous expectation of AI-driven layoffs among global CEOs is not currently matched by evidence of aggregate job destruction in the most rigorous empirical studies available. Instead, the labor market is undergoing a significant compositional shift, reducing demand for structured, repetitive work while increasing it for analytical, creative, and collaborative roles. The heaviest near-term burdens fall on workers in automation-exposed entry-level positions and on the broader workforce's psychological well-being, where anxiety about job insecurity is already taking a measurable toll. For workers and organizations alike, the challenge is not simply surviving a wave of mass layoffs—it is navigating a fundamental redefinition of which skills the market values and how quickly people can adapt.