Meta is reportedly planning layoffs that could affect up to 20% of its workforce. This would equate to roughly 16,000 employees. Three anonymous sources told Reuters the plans are underway, driven by soaring AI infrastructure costs and a push for efficiency. No final decision on timing or scale has been made, and Meta called the reports speculative.
Roughly 16,000 of its nearly 79,000 employees could be out of a job if the claims are correct. Three anonymous sources familiar with the matter told Reuters about the plans. No date has been set and the final scale has not been decided. Meta spokesman Andy Stone called it “speculative reporting about theoretical approaches.”
If confirmed, this would be Meta’s largest restructuring since the company’s “year of efficiency” in late 2022 and early 2023, when it cut approximately 21,000 jobs in two rounds. During that time, several Big Tech companies opted to execute similar workforce cuts, including Google, which until January 2023 had not felt the need to slash jobs like its tech rivals.
From metaverse to superintelligence
The planned cuts follow an earlier, smaller round of job losses. In January, Meta laid off around 10 percent of its Reality Labs division, which focuses on VR and metaverse projects. The latest round would be far more sweeping and company-wide. The earlier workforce cut could easily be explained by the fact that the metaverse has seemingly vanished from the popular vocabulary around the same time ChatGPT took the world by storm, and AI as both a technology and a buzzword with it.
The new push for efficiency is directly linked to Meta’s soaring AI spending. The company has budgeted up to 135 billion dollars for 2026 capital expenditures, nearly double the $72 billion spent in 2025, with a 600 billion dollar data center construction programme planned through 2028. It has also offered pay packages worth hundreds of millions of dollars to recruit AI researchers, and recently acquired Moltbook, a social network for AI agents. Meta is also spending at least 2 billion dollars on Chinese AI startup Manus.
AI models lagging behind
Despite the investment push, Meta’s AI efforts have hit obstacles. Avocado, the model being built by its superintelligence team, has been delayed until at least May and does not yet match top models from Google, OpenAI, and Anthropic. Last year, Meta’s Llama 4 Behemoth model was shelved after missing benchmarks. Despite positive benchmarks in AI models before that cancelled release, Llama LLMs have also failed to retain the open-source community it had built up with the releases of Llama 2 and Llama 3.
Zuckerberg said in January he was starting to see “projects that used to require big teams now be accomplished by a single very talented person.” However, this framing needs scrutinizing. Other large tech companies with enormous planned AI buildouts have needed to prioritize the AI infrastructure spend itself, such as Oracle and Amazon, with the threat of large-scale layoffs looming.
Amazon cut 16,000 jobs in January, also citing AI-driven efficiency, where one can assume that the AI factor here may also relate to opening up funds for future data centers. Nevertheless, Block recently even halved its workforce, with CEO Jack Dorsey explicitly pointing to AI tools and their growing capability to help companies do more with smaller teams. Again, though, other factors are at play too, and Dorsey’s alleged over-hiring at Twitter supposedly required large-scale cuts when Elon Musk took over the company and rebranded it to X.