The number on Layoffs.fyi ticked past 128,270 in late May 2026, each digit representing a person who opened a calendar invite or a Slack message and learned their job no longer existed. Engineers at Microsoft, which confirmed a performance-focused reduction spanning thousands of roles across divisions. Product managers at SAP, where leadership tied a sweeping restructuring to the company’s AI pivot. Entire teams at smaller startups that quietly dissolved between funding rounds, their Slack workspaces going dark without a press release.
In nearly every case, executives pointed to artificial intelligence as the catalyst. The cuts were framed as necessary to “reallocate resources toward AI” or “streamline operations through automation.” But two major academic studies now challenge the core assumption behind those decisions: that swapping human workers for AI tools reliably makes a company more productive. The research suggests the odds of disruption are roughly equal to the odds of efficiency gains, a finding that should give pause to any board treating headcount reduction as a shortcut to growth.
What the layoff tracker reveals about 2026
Layoffs.fyi started as a side project by tech recruiter Roger Lee during the pandemic and was later profiled by The New York Times. It aggregates publicly announced layoffs at technology and startup companies, pulling from press releases, WARN Act filings, leaked internal memos, and verified media reports. It does not capture quiet attrition or contractor cuts, so the real number is almost certainly higher.
The 2026 pace stands out less for its raw scale than for its stated rationale. The 2023 tech downturn, which The Wall Street Journal reported was accelerating faster than at any point during the pandemic, ultimately produced higher totals across the full year. But those cuts were driven by macroeconomic anxiety and post-pandemic overcorrection. In 2026, company after company has invoked AI transformation as the primary justification, turning what used to be a cost-cutting story into a technology story with very different implications for the workers involved.
What two major studies actually found
The first study, National Bureau of Economic Research Working Paper No. 34984, titled “Artificial Intelligence, Productivity, and the Workforce: Evidence from Corporate Executives,” surveyed hundreds of corporate leaders about the real-world results of AI adoption. The findings were decidedly mixed. Executives reported efficiency improvements in specific functions, particularly customer service automation and code generation. But they reported disruption, retraining costs, and coordination breakdowns at comparable rates. The paper’s central conclusion: net productivity gains from AI were far from guaranteed, and the variance across firms was enormous.
The second study, a Stanford Institute for Economic Policy Research working paper titled “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence,” took a different approach. Researchers used high-frequency administrative payroll data from a large U.S. payroll provider to track employment shifts in occupations most exposed to AI. Their central finding: early-career workers in AI-exposed roles experienced a measurable decline in employment, but there was no corresponding surge in hiring, revenue growth, or productivity elsewhere in the same firms. Managers appeared to be trimming junior positions as automation tools arrived while holding on to experienced staff who could supervise or complement the technology.
Neither paper claims AI is useless. Both document real pockets of efficiency. But the pattern they describe is one where companies that cut staff in the name of AI are roughly as likely to absorb transition costs, lose institutional knowledge, and face coordination problems as they are to pocket clean savings.
The gaps that make certainty impossible
Connecting the 128,270 layoffs directly to AI replacement decisions is harder than it looks. The U.S. Bureau of Labor Statistics does not isolate AI-driven job cuts in its monthly reports, so no official government dataset confirms what share of these losses stem from automation versus garden-variety restructuring, offshoring, or project cancellations.
Company disclosures offer little clarity. Earnings calls and SEC filings from major tech employers lean on vague language: “efficiency,” “organizational realignment,” “sharpening our focus.” Rarely does a filing specify that a particular team was replaced by an AI system. That ambiguity makes it difficult to separate firms that genuinely automated work from those that borrowed AI rhetoric to justify cost cuts they would have made regardless.
The two academic studies, while rigorous, carry their own limitations. The NBER paper relies on executive self-reporting, which can skew optimistic. The Stanford paper captures a snapshot of payroll trends but does not follow individual companies from the moment they adopt AI through a full business cycle. Without that longitudinal view, it is impossible to know whether early productivity bumps persist, flatten, or reverse as organizations adapt.
The Stanford researchers are candid about this. They frame their results as early signals, not settled conclusions, and warn that the employment decline among early-career workers could represent either a temporary hiring freeze or the beginning of a more structural divide between workers whose skills complement AI and those whose tasks are largely automated.
Why the “do more with less” playbook keeps failing
If the research is right that AI-driven headcount cuts produce uneven results, the obvious question is why so many companies keep making them.
Financial signaling is one reason. Announcing layoffs alongside an “AI transformation” strategy tends to boost share prices in the short term, regardless of whether the transformation delivers. Investors reward the narrative of efficiency even before the efficiency materializes, creating an incentive structure that favors cuts over patience.
Competitive pressure is another. When one major employer announces it is replacing a function with AI, rivals feel compelled to follow or risk looking slow. That dynamic can produce industry-wide cuts that outpace the technology’s actual readiness. The NBER paper’s authors flag this pattern, noting that coordination problems and retraining burdens often surprised executives who expected smoother rollouts.
Then there is the simplest explanation: cutting payroll is faster than building new revenue. AI tools can reduce the cost of certain tasks within weeks. Building the organizational capacity to use those tools well, retraining remaining staff, redesigning workflows, maintaining quality without the institutional knowledge that departed employees carried, takes months or years. The Stanford study’s finding that firms did not see offsetting growth after trimming junior roles fits this lag: the savings showed up immediately, but the supposed productivity gains had not followed by the time the researchers measured.
Who is most exposed, and what comes next
The data do not yet show a wholesale collapse of tech employment. The sector still pays well above the national median, and many laid-off workers have found new roles, often at companies that are themselves hiring for AI-related positions. But the research points to a specific and growing risk: early-career employees in routine, digitally mediated roles are the most vulnerable when employers adopt AI tools quickly. That group has the least bargaining power and the fewest savings to fall back on during a job search.
For executives and boards, the message from both studies is sobering. Cutting headcount in the name of automation does not guarantee durable efficiency. The transition costs, coordination failures, and loss of institutional memory documented in the research can offset or even outweigh the promised savings. Companies that treated earlier rounds of layoffs as a quick fix often struggled to rebuild capacity when demand returned, and there is no reason to assume this cycle will play out differently.
Until more granular, long-term data are available, sweeping claims that AI is either definitively destroying jobs or unquestionably boosting productivity will overstate what the evidence supports. What the numbers and the research do confirm is narrower but still significant: the current wave of AI-driven layoffs is real, the corporate logic behind it is shakier than the press releases suggest, and the people absorbing the most risk are the ones least equipped to carry it.
Warren Cohen is a finance writer based in Phoenix, Arizona, covering personal finance topics including credit, banking, and beginner investing. He earned his degree in business administration from Arizona State University and began his career working in consumer finance, where he gained direct experience with lending and credit systems. He now writes for personal finance websites and fintech platforms, focusing on clear, practical content that helps readers make informed financial decisions.
