Meta tech layoffs on a scale the company has never attempted before are reportedly in the works, according to reports that emerged on March 14, 2026. With nearly 79,000 employees on payroll as of December 31, 2025, a 20% cut would eliminate somewhere between 15,800 and 16,000 jobs. No date has been set, the final scope is unconfirmed, and Meta spokesperson Andy Stone has called the reports “speculative reporting about theoretical approaches”. That caveat matters. But the direction of travel is hard to ignore.
TL;DR: Meta is reportedly weighing layoffs that could affect 20% of its workforce — potentially 15,800-16,000 people — driven by surging AI infrastructure costs, massive pay packages for AI talent, and a push to replace team-heavy workflows with AI-assisted ones. No date or confirmed scale yet.
Why Meta tech layoffs are happening now
The core driver is straightforward: AI is expensive, and Meta is paying for it twice. First, through the infrastructure — data centres, chips, energy — needed to train and run large models. Second, through the talent, with top AI researchers reportedly receiving pay packages worth hundreds of millions over four years. Something has to give, and headcount is the lever being pulled.
Mark Zuckerberg framed the logic bluntly in January, arguing that “projects that used to require big teams now be accomplished by a single very talented person”. That is not a throwaway line. It is the ideological scaffolding for what could be the largest single workforce reduction in Meta’s history — bigger than the 11,000 jobs cut in November 2022 and the further 10,000 eliminated in 2023. Top executives have already been instructed to prepare cost-cutting strategies, signalling that internal planning is real even if the public stance remains dismissive.
Meta’s AI models are struggling, and that makes the cuts harder to justify
The uncomfortable context here is that Meta’s AI ambitions have hit visible turbulence. The Llama 4 models drew criticism for misleading benchmark results, and the largest planned version, codenamed Behemoth, was abandoned after originally targeting a summer 2025 release. A newer model internally codenamed Avocado has slipped from a March target to at least May 2026, and internal tests reportedly show it underperforming rivals including Google’s Gemini 3.0 in reasoning, coding, and writing.
Meta is not standing still. The company has stood up a Superintelligence team and an elite unit called TBD Lab, led by Wang, which released the Vibes AI video app. These are real bets. But cutting thousands of workers while your flagship AI models are delayed and underperforming is a difficult story to tell — internally or externally. The efficiency argument only lands if the AI actually works.
Is this an industry trend or just AI-washing old pandemic over-hiring?
Meta is not alone. Amazon carried out 16,000 job cuts in January 2026, following earlier rounds of roughly 30,000 white-collar reductions, with AI efficiency cited as a driving rationale. Block, the payments company led by Jack Dorsey, went further still, laying off nearly half its workforce. Dorsey’s stated reasoning: “AI tools would allow companies to do more work with fewer people”. The pattern is consistent enough across major tech firms that it reads as a genuine structural shift, not a series of isolated decisions.
That said, some observers have noted that Big Tech massively over-hired during the pandemic era, and that AI is being used as a convenient narrative to justify corrections that were coming anyway. OpenAI’s Sam Altman has made similar observations publicly. Both things can be true simultaneously: companies genuinely are reorganising around AI productivity, and they are also tidying up hiring decisions made in a very different economic moment. The workers being let go are real either way.
What does this mean for Meta’s workforce going forward?
Meta’s recent layoffs in January 2026 already cut 1,500 positions at Reality Labs, the division responsible for VR and AR hardware. That unit has been a persistent money sink, and trimming it while doubling down on AI reflects a clear strategic priority shift. The reported 20% figure, if it materialises, would represent a company that looks fundamentally different by the end of 2026 — smaller, more AI-dependent, and betting heavily that the models currently underperforming in internal tests will eventually catch up to the ambition.
How many jobs could Meta cut in 2026?
Based on Meta’s headcount of nearly 79,000 employees as of December 31, 2025, a 20% reduction would mean roughly 15,800 to 16,000 job losses. That would surpass the 11,000 cuts made in November 2022 and make it the largest single round of layoffs in the company’s history. No timing or final number has been confirmed by Meta.
Why is Meta cutting jobs if it is investing in AI?
Meta is cutting jobs precisely because of its AI investment, not despite it. The cost of AI infrastructure and the pay packages required to attract top AI researchers are straining the budget, so the company is reducing headcount elsewhere to offset those expenses. Zuckerberg has also argued that AI tools allow smaller teams to accomplish what previously required much larger ones, which changes the calculus on how many people Meta needs at all.
How do Meta’s layoffs compare to other Big Tech cuts?
Amazon cut around 16,000 jobs in January 2026 following earlier rounds of roughly 30,000 white-collar reductions, both linked to AI efficiency arguments. Block went further proportionally, cutting close to half its workforce under Jack Dorsey’s AI productivity rationale. Meta’s reported 20% cut, if confirmed, would be larger in absolute numbers than either Amazon round and would signal that no corner of Big Tech is immune to this particular wave of restructuring.
Meta’s denials should be taken seriously — these reports are based on anonymous sources and the company has explicitly pushed back. But the structural pressures are real, the internal planning signals are real, and the industry context is real. If the Meta tech layoffs do materialise at the scale being reported, they will mark a defining moment in how the AI era reshapes the companies that are supposedly building it.
This article was written with AI assistance and editorially reviewed.
Source: TechRadar
