Meta’s AI spending spree could trigger its biggest layoffs yet

The Facebook and Instagram parent is reportedly weighing sweeping job cuts even as its ad machine keeps printing money, a sign that Silicon Valley’s AI race is turning into a brutal exercise in cost discipline as much as innovation.

Meta may be headed for its most dramatic workforce reduction since Mark Zuckerberg’s so-called “year of efficiency.” Reuters reported that the company is discussing layoffs that could affect 20% or more of its workforce, though no date has been set and the final scope has not been decided. Meta spokesperson Andy Stone pushed back, calling it “speculative reporting about theoretical approaches.” Still, even the possibility is striking: Meta reported 78,865 employees at the end of 2025, meaning a 20% cut would land in the neighborhood of 15,700 to 15,800 jobs.

What makes the story more jarring is that this is not a company visibly in financial distress. Meta reported $200.97 billion in 2025 revenue, up 22% year over year, with $72.22 billion in capital expenditures last year alone. It finished 2025 with $81.59 billion in cash, cash equivalents, and marketable securities. In other words, this is not a survival layoff story in the old recession-era sense. It is a capital-allocation story, where a highly profitable company is trying to redirect staggering amounts of money toward artificial intelligence infrastructure.

And the amounts are staggering. Meta told investors in January that it expects 2026 total expenses of $162 billion to $169 billion and 2026 capital expenditures of $115 billion to $135 billion, with the biggest increase driven by infrastructure costs. The company said that spending is meant to support its Meta Superintelligence Labs efforts as well as its core business. Reuters separately reported that Meta’s broader data-center buildout could total $600 billion by 2028, a figure that helps explain why executives may be looking for labor savings elsewhere in the system.

That is the central irony of the moment. Meta’s ad business remains powerful, but the cost of staying competitive in AI is ballooning so quickly that even a company with huge revenue and cash reserves appears willing to consider another round of cuts. Reuters reported that Zuckerberg has been offering enormous compensation packages to recruit top AI researchers, acquired the AI-agent-focused platform Moltbook, and is spending at least $2 billion to buy Chinese startup Manus. At the same time, the company has been trying to convince itself and investors that better AI tools can allow smaller teams to do work that used to require far more people.

There is also a performance anxiety humming beneath the budget math. Reuters said Meta’s recent AI push follows setbacks with its Llama 4 models, including criticism over benchmark presentation, the abandoned release of its largest planned version called Behemoth, and underwhelming progress on a newer model called Avocado. That matters because it reframes the layoffs discussion. This is not simply about trimming fat from a mature tech giant. It is also about trying to buy time and compute power in an AI race where falling behind is treated almost like a strategic emergency.

Wall Street, predictably, liked the sound of discipline. Reuters reported that Meta shares rose about 3% in premarket trading on March 16 after the layoffs report, making the company one of the day’s top gainers before the bell. That reaction says a lot about the era. Investors appear willing to reward companies for cutting people if those cuts help fund AI infrastructure or signal that AI may eventually reduce labor intensity. In that sense, Meta’s reported plans are not an outlier. Reuters noted that Amazon and Block have already made similar AI-linked workforce cuts this year.

The bigger business story, then, is not just that Meta may cut jobs. It is that AI is reshaping corporate priorities in a way that makes even booming companies act leaner, colder, and more impatient. The old tech script was growth first, profits later. The new one looks more like this: spend ferociously on chips, data centers, and elite researchers, then demand that the rest of the organization shrink to make the numbers behave. Meta may still decide against cuts on the scale Reuters reported. But the fact that such a move is under discussion at one of the richest companies in the world says plenty about where this cycle is heading.