As company pours billions into AI, insiders brace for potential job losses while company’s share prices rise
Meta Platforms is said to be preparing layoffs affecting as much as a fifth of its workforce.
Shares rose on the news even as Meta publicly pushed back on suggestions it has already decided on sweeping cuts.
According to an exclusive report from Reuters, Meta is exploring plans for layoffs that could hit 20% or more of its staff, or roughly 16,000 employees, as it looks to offset surging AI infrastructure costs and reshape how work is done inside the company. Three people familiar with the discussions told Reuters that internal planning is under way, though the precise timing and final scale of any cuts have not been settled.
CNBC reported that Meta’s stock climbed in premarket trading on Monday after the Reuters story, with investors betting that a leaner cost base could bolster profits even as the company doubles down on AI.
Shares climbed nearly 3% in early trading, extending a multi‑year rally that has been fueled in part by its earlier layoffs and spending discipline.
Layoffs: ‘Theoretical approaches’?
Publicly, though, Meta is trying to tamp down speculation about a new wave of job losses. The company told Reuters that discussions cited in the report referred to options rather than firm decisions, and Meta spokesperson Andy Stone described it as “speculative reporting about theoretical approaches.”
Techzine reported that a reduction on the scale described by Reuters would represent Meta’s largest restructuring since its self‑proclaimed “year of efficiency” in late 2022 and early 2023, when it eliminated around 21,000 roles across two major rounds of layoffs. Those cuts, mirrored across Silicon Valley by companies including Google and Amazon, marked one of the deepest contractions at a major U.S. technology firm since the dot‑com bust.
The latest restructuring discussions come just months after a smaller but notable round of job losses inside Reality Labs, the unit responsible for virtual reality hardware and metaverse initiatives. In January, Meta laid off about 10% of that division, Techzine reported, as the company shifted attention and resources away from the metaverse push that dominated its strategy only a few years ago.
Two in three employers that reduced headcount because of artificial intelligence are already rehiring laid off staff, as most express regret over how they handled the AI-led retrenchments, found a recent survey.
AI spending on unprecedented scale
Behind the new round of belt‑tightening is an AI spending plan of unprecedented scale. Reuters reported that Meta has budgeted as much as $135 billion in capital expenditures for 2026 alone, nearly twice the $72 billion it spent in 2025, tied largely to data centers, specialized chips and other infrastructure to train and run large AI models.
Over the longer term, Meta is preparing for a data center build‑out that could total roughly $600 billion through 2028, according to details cited by Reuters.
The company has also poured hundreds of millions of dollars into compensation packages for top AI researchers and engineers, and has been on an acquisition push that includes the recent purchase of Moltbook, a social network for AI agents, Techzine reported. On top of that, Meta is spending at least $2 billion on Chinese AI startup Manus, further intensifying the budget pressure around its AI strategy.
Delays to AI model by Meta
Yet the payoff from those investments has been uneven so far. Techzine, drawing on Reuters’ reporting, said Meta’s flagship new model — known internally as Avocado and being developed by its “superintelligence” team — has been delayed until at least May and still lags leading systems from OpenAI, Google and Anthropic.
A previous attempt at a breakthrough system, the Llama 4 “Behemoth” model, was shelved after failing to meet performance targets, according to the same report.
Meta has positioned its Llama family of models as a pillar of open‑source AI, but Techzine noted that enthusiasm among developers has cooled compared with the release of Llama 2 and Llama 3, even as rivals have accelerated.
AI cuts: From big teams to one person
Inside the company, executives have sought to frame AI as a tool that will eventually allow smaller teams to do more. In January, CEO Mark Zuckerberg said he was starting to see “projects that used to require big teams now be accomplished by a single very talented person.”
That narrative aligns with a broader message across Big Tech that automation will raise productivity, but it also fuels anxiety among employees that AI will be used to justify leaner headcounts.
Other tech giants are already drawing a clearer line between AI investments and job cuts. Techzine pointed to Amazon, which cut 16,000 jobs in January, with executives citing AI‑driven efficiency as a factor. Payments company Block, led by Jack Dorsey, has also slashed its workforce roughly in half, explicitly linking the move to AI tools that it says enable the company to operate with fewer people, according to Techzine.
