Meta Trims 600 Roles in AI Push, Says Smaller Teams Will Move Faster

Axios, CNBC, the New York Times, Reuters, Business Insider, and the Verge contributed to this report.

Meta is cutting about 600 jobs across its artificial-intelligence group, the company confirmed, as it tries to strip out bureaucracy and speed up decision-making while still spending big on AI.

The layoffs, announced in a memo from chief AI officer Alexandr Wang, hit parts of FAIR (Facebook AI Research), product-facing AI and AI infrastructure. The newly created TBD Lab — the elite team training Meta’s next-gen foundation models — won’t be touched and is still hiring, according to the memo.

“By reducing the size of our team, fewer conversations will be required to make a decision, and each person will be more load-bearing,” Wang told staff.

US employees were notified Wednesday morning; some were placed on a non-working notice period through Nov. 21 and can apply for other roles inside Meta. Severance includes 16 weeks of pay plus two weeks per year of service (minus the notice period), per internal messages viewed by CNBC.

The cuts land after a frenetic year of AI spending and reshuffling. Meta poured billions into infrastructure and talent to keep pace with OpenAI and Google, invested $14.3 billion in Scale AI, hired Wang in June, and stood up “Meta Superintelligence Labs” to concentrate its top researchers and engineers. Despite that spree, CEO Mark Zuckerberg has been unhappy with the pace — especially after April’s Llama 4 release drew a lukewarm developer response — prompting a pivot to “small, talent-dense teams.”

Even as jobs are trimmed, the company is doubling down on the plumbing required for ever-larger models. On Tuesday, Meta unveiled a $27 billion financing deal with Blue Owl Capital to fund its Hyperion data center project in Louisiana — huge enough, Zuckerberg has said, to cover a “significant part of the footprint of Manhattan.”

Meta reports Q3 results next week, with expenses already guided higher for 2025 and management warning that AI spending will push 2026 costs up even more. The message to investors and employees alike: fewer layers, faster ships — but the AI bet isn’t getting smaller.