Meta Platforms Inc. will lay off approximately 600 employees from its Superintelligence Labs division, as the tech giant undertakes a major restructuring to streamline its artificial intelligence operations and refocus on high-priority initiatives.
The cuts, disclosed in an internal memo from Meta’s Chief AI Officer Alexandr Wang, impact several legacy teams, including the Facebook AI Research (FAIR) group, product-focused AI units, and infrastructure support teams. Notably, the newly formed “TBD Lab”—tasked with advancing Meta’s long-term “superintelligence” ambitions—has been spared.
The reductions affect a small fraction of Meta’s several-thousand-strong AI workforce.
U.S.-based employees received notifications as early as 7 a.m. on October 22, with international staff informed over the following days.
Meta is actively encouraging impacted workers to apply for other internal roles, with the company expecting most to transition elsewhere within the organization.
The move follows a turbulent year for Meta’s AI division, including the controversial rollout of its open-source Llama 4 model, which drew criticism for underwhelming performance relative to expectations. The model’s reception reportedly triggered internal scrutiny and a leadership reshuffle in June 2025.
Despite the job cuts, Meta continues to invest heavily in AI. The company poured $14.3 billion into Scale AI—Wang’s former company—earlier this year, and offered more than $100 million in bonuses to poach top talent from rivals such as OpenAI and Google DeepMind. Meta has also committed $27 billion in AI financing for 2025, even as it paused broader AI hiring earlier this year.
Industry analysts view the restructuring as a sign of mounting pressure to convert AI research into real-world, revenue-generating products. Meta is pivoting from broad-based research efforts toward smaller, product-driven teams, mirroring a broader trend across Big Tech.
As Meta sharpens its focus on “superintelligence,” the stakes in the AI arms race are only growing. With competition from OpenAI, Anthropic, Google, and a surge of well-funded startups, the coming year is likely to test whether Meta’s new strategy can deliver both innovation and impact.