The running list: Major tech layoffs in 2026 where employers cited AI
Source: TechCrunch
Published: 2026-07-25
Entity Analyzed: Tech Capital Reallocation
URL SCAN
TechCrunch’s updated running list of 2026 tech layoffs leads with Monday.com’s 20% cut (~630 employees) for an ‘AI Work Platform’ pivot, while a Financial Times analysis reveals that companies citing AI in layoffs have underperformed the Nasdaq by ~10% post-announcement. Nearly 140,000 tech jobs have been cut this year, with Amazon, Oracle, Meta, and Microsoft alone accounting for ~50,000.
The Triage
This article is doing something the genre rarely does: it includes the data that undermines its own frame. The running list format — cataloging layoff after layoff, company after company — is designed to convey inevitability. But buried in the fourth paragraph is a Financial Times analysis that detonates the narrative: companies that cite AI as a factor in layoffs underperform the Nasdaq by approximately 10% in the 30 trading days after the announcement. The market is not buying the story. Investors are not rewarding AI-driven restructuring; they are penalizing it. The layoffs are not a signal of technological leadership. They are a signal of managerial desperation dressed in Silicon Valley vernacular.
The Monday.com opening is the perfect case study. Eran Zinman’s memo — ‘not made to reduce costs or replace people with AI’ — is a masterclass in corporate doublespeak. The company is firing 20% of its staff, booking $45-55 million in restructuring charges, and pivoting to an ‘AI Work Platform’ while insisting the two are unrelated. The framing is not adaptation; it is alibi. The ‘AI-first vision’ is the press release; the headcount reduction is the balance sheet. That the company still projects 20% YoY revenue growth only deepens the contradiction: if the business is healthy and growing, why does it need to fire 630 people to ‘adapt’ to a technology it claims is not replacing them?
The running list itself is the autopsy. Nineteen companies. Nineteen admissions. Nearly 140,000 jobs eliminated in a single year, with the four largest tech employers alone accounting for 50,000. The scale is not a trend; it is a sector-wide liquidation of the employment contract that built the industry. And the most telling detail is not in the layoffs but in the hires: Anthropic and OpenAI are ‘hiring rapidly,’ while IBM is tripling entry-level AI hiring. The jobs are not vanishing; they are migrating — from established tech companies with legacy workforces to AI-native companies with tiny headcounts and massive compute budgets. The worker is not being replaced by AI. The worker is being replaced by a different company that uses AI.
The Autopsy (with DT-LAG)
Mechanical Collapse Point
The mechanical collapse is the decoupling of layoff decisions from market performance. The FT analysis is the critical signal: when companies announce AI-driven layoffs, their stock underperforms by 10%. This means the layoffs are not investor-driven efficiency moves; they are managerial gambits that the market actively distrusts. The mechanical driver is not shareholder pressure for higher margins. It is competitive panic: when Oracle fires 21,000 people ‘to invest in AI,’ every other company in the sector must either match the signal or be perceived as falling behind. The result is a coordination failure where everyone fires because everyone else is firing, and the market punishes everyone for it.
The specific company data reveals the mechanics of this panic. Meta: 7,000 employees moved into AI-focused roles while 8,000 are laid off. This is not restructuring; it is workforce arbitrage — replacing expensive generalists with cheaper AI specialists, or more precisely, replacing humans who build products with humans who tune models. IBM: tripling entry-level AI hiring while simultaneously cutting 3,000-9,000 U.S. positions. The tripling is not growth; it is replacement at lower cost. Entry-level AI roles pay less than the senior operational roles being eliminated. The ‘AI and hybrid-cloud’ roles that IBM is hiring for are not new jobs; they are the same jobs, rebranded and repriced.
The Monday.com case is the most mechanically transparent. $45-55 million in restructuring charges to fire 630 people, while projecting 20% revenue growth. The math does not work as cost-cutting; it works as narrative investment. The company is spending $45-55 million to buy the ‘AI-first’ positioning that investors are supposed to reward — except the FT data shows they don’t. The restructuring charge is the price of admission to a narrative that the market has already stopped believing.
Lag-Weighted Social Timeline
The lag is 3-6 months for narrative collapse, 12-18 months for workforce recognition, and 2-4 years for institutional adaptation. In the immediate term — Q3-Q4 2026 — the ‘AI layoffs’ narrative will continue to dominate because it serves a critical political function: it allows companies to fire people without admitting that the firings are not generating returns. The FT underperformance data will circulate in investor circles but will not break into mainstream coverage because it undermines the story that both executives and journalists need to tell.
By early 2027, the first wave of ‘AI regret’ will manifest as earnings crises. Companies that spent billions on AI infrastructure while firing their revenue-generating workforces will face the moment when their compute bills are due and their remaining staff is too small to execute. The article’s observation that AI-focused companies are hiring rapidly is the canary: the talent is not leaving tech; it is leaving legacy tech for AI-native companies. The ‘brain drain’ from Microsoft, Oracle, and Amazon to Anthropic and OpenAI is not a migration of skills; it is a migration of the industry’s future away from the companies that built it.
By 2028-2029, the political and regulatory response will coalesce around the capital reallocation documented in the article. The ‘payroll to compute’ shift — with companies simultaneously firing workers and building data centers — is not just a labor story; it is a tax-base and community story. When 140,000 tech workers lose their jobs in a single year, the housing markets, service economies, and municipal budgets of tech hubs experience cascading stress. Meanwhile, the data centers being built are often in jurisdictions with tax incentives, meaning the social cost is concentrated in San Francisco and Seattle while the infrastructure benefits are distributed to states and countries competing for AI investment. This geographic mismatch will drive political backlash that targets not AI itself but the corporate governance model that uses AI as justification for workforce liquidation.
Lag Factors
Narrative Momentum vs. Market Signal: The FT underperformance data is a lag factor because it is known to investors but invisible to the public. The gap between what the market knows (AI layoffs are punished) and what companies do (continue announcing them) creates a self-reinforcing cycle of bad decisions. The lag is the time it takes for this gap to close.
Competitive Panic Cascade: When one company in a sector announces AI layoffs, the others must follow or be perceived as behind. The article’s running list is itself a lag accelerator: by cataloging the cuts, it creates a reference point that board members use to pressure their own CEOs. The lag is the time it takes for this cascade to exhaust itself.
Infrastructure Lock-in: The hundreds of billions committed to AI data centers are not reversible. Once a company has signed the power contracts and purchased the GPUs, the economic pressure to justify the investment through workforce reduction becomes overwhelming — even when the market signals that the strategy is failing. The lag is the time between infrastructure commitment and the inevitable layoffs that follow.
Skill Migration Asymmetry: The article notes that Anthropic and OpenAI are hiring rapidly while legacy companies fire. This creates a two-tier labor market where the best talent migrates to AI-native companies, leaving legacy companies with a degraded workforce that is more expensive to replace than the AI that supposedly replaces it. The lag is the time it takes for legacy companies to realize they have fired the people who understood their systems.
Stock Option Golden Handcuffs: Remaining employees on vesting schedules delay their departure, creating a false sense of workforce stability. The lag is the time between vesting and the exit of institutional knowledge.
Pandemic Overhang Confound: The article does not explicitly mention the pandemic hiring bubble, but many of the companies on the list (Microsoft, Amazon, Meta) over-hired in 2020-2021. The 2026 layoffs are partly a correction of that over-hiring, but the ‘AI restructuring’ label prevents the social recognition that the layoffs are partly historical normalization.
Defensive Moats
Regulatory Armor: The article notes that AI-focused companies are hiring rapidly, suggesting that the ‘AI jobs’ category is not disappearing but relocating. This relocation creates a temporary moat for workers in regulated industries (healthcare, finance, government) where AI-native companies have not yet penetrated. But the moat is shallow: the running list already includes Coinbase, PayPal, and Standard Chartered (from related coverage), indicating that financial services are not insulated.
Trust Shield: The ‘human judgment’ argument is being actively undermined by the companies themselves. When Monday.com’s co-founder says AI is not replacing humans while firing 20% of his staff, the trust shield is not a moat; it is a contradiction. The market’s 10% underperformance penalty suggests that investors, at least, are not fooled.
Physical Chains: The data centers being built require human construction, maintenance, and security. But the article’s focus on the running list of layoffs misses the human labor that still surrounds the compute. The lag is the time it takes for companies to realize that the data center needs humans too — and that those humans are not cheaper than the software engineers they fired.
Future-Proofing Scorecard
| Timeline | Score | Commentary |
|———-|——-|————|
| 1 year | 2/10 | The ‘AI layoffs’ narrative persists despite market underperformance. Layoffs continue at current pace. AI-native companies absorb talent from legacy companies. First wave of ‘AI regret’ hits over-leveraged infrastructure spenders. |
| 2 years | 1/10 | The divergence between AI-native and legacy tech becomes a chasm. Legacy companies face talent shortages, maintenance crises, and declining market share. The ‘running list’ of layoffs is replaced by a ‘running list’ of restructurings and acquisitions. |
| 5 years | 0/10 | The tech industry has bifurcated: AI-native companies with sub-1,000-person teams and massive compute budgets, and legacy companies that either transformed or died. The concept of ‘tech employment’ as a stable career path is obsolete for all but the elite few. |
| 10 years | 0/10 | The 140,000 figure is a historical footnote. The ‘tech worker’ category has been replaced by ‘compute operator’ and ‘AI systems architect.’ Human labor in technology exists only in regulatory-mandated oversight, physical infrastructure maintenance, and elite research. The employment contract that built Silicon Valley is a memory. |
The Verdict
The TechCrunch running list is a ledger of a sector in the grip of a collective delusion. The delusion is that firing people and buying chips is a growth strategy. The Financial Times analysis — that AI-citing companies underperform by 10% — is the market’s verdict, delivered in the only language executives understand: stock price. And the market is saying no. The market is saying that ‘AI restructuring’ is not a signal of technological leadership; it is a signal of managerial failure, of companies that do not know how to grow their top line and have resorted to shrinking their cost base while pretending it is innovation.
Monday.com’s Eran Zinman, with his $45-55 million restructuring charge and his insistence that AI is not replacing humans, is the archetype of this delusion. He is spending shareholder money to buy a narrative that the market has already rejected. The 630 people he fired are not being replaced by AI; they are being sacrificed to a positioning statement. And the 140,000+ tech workers cut this year are not being replaced by AI either; they are being liquidated to fund infrastructure that may or may not deliver returns, by companies that have forgotten how to grow revenue and have settled for cutting costs.
The verdict: this is not the story of AI replacing workers. It is the story of an industry that has lost its way, firing the people who built it to fund a technological bet that the market does not believe in. The running list will grow. The underperformance will continue. And the only winners will be the AI-native companies — Anthropic, OpenAI, and their ilk — that are hiring the displaced talent at depressed wages while the legacy companies that fired them wonder why their products no longer work. The 10% penalty is not a discount. It is a warning. And the industry is not listening.