Meta has let go of around 600 employees from its artificial intelligence department in a major restructuring as the social media giant tries to re-streamline and reboot its AI ambitions with new leadership.
The company announced the reductions on Wednesday, impacting employees in Meta’s AI infrastructure groups, its FAIR unit of fundamental artificial intelligence research, and different product-related roles.
The move follows as Meta continues to spend billions of dollars on artificial intelligence at the same time as it is attempting to weed out what company insiders have characterized as organizational bloat.
The reductions reflect a clear priority by CEO Mark Zuckerberg for his pricey new recruits over existing employees. Most significantly, employees at TBD Labs, which houses several of the top AI researchers hired this summer- were entirely exempt from the reductions. They report to Alexandr Wang, who arrived at Meta in June as chief AI officer after the company’s whopping $14.3 billion purchase of Scale AI.
This contrast highlights Zuckerberg’s gamble on new talent to inject life into Meta’s AI initiative. Those close to the situation informed CNBC that the AI division had swelled in size over time, with FAIR and product-driven groups often vying for computer resources.
When Meta’s new recruits showed up to establish Superintelligence Labs, they took over an oversized entity that required streamlining.
Meta Cuts Superintelligence Labs Staff Amid AI Frustration
Since the layoffs, Meta’s Superintelligence Labs has only around 3,000 employees. Those affected were informed on Wednesday that they will be let go on November 21, although they’ve been put on a “non-working notice period” from today.
“Through this period, your internal access will be revoked, and you don’t have to perform any further work for Meta,” the company informed, leaving employees in a message seen by CNBC. “You can use this time to look for another position at Meta.”
Meta is providing a 16-week severance pay plus two weeks for each full year of service, less the notice period. Meta has also provided the impacted workers with a chance to apply for other roles within Meta before leaving officially.
The reorganization is while increasing frustration from Zuckerberg regarding Meta’s progress on AI. The firm’s Llama 4 versions, launched in April, saw lukewarm uptake from developers, compelling the CEO to act forcefully.
This frustration contributed to Meta’s keenness on heavily investing in Scale AI and building Meta Superintelligence Labs, a new unit headed by Wang and former GitHub CEO Nat Friedman.
The division unites the world’s best AI researchers and engineers with the very specific mission of assisting Meta in keeping up with the competition in the form of OpenAI and Google, which have been leaders in the advancement of artificial intelligence.
Meta Doubles Down on AI Investment Despite Layoffs, Escalating Expense Guidance
Even with the layoffs, Meta has no indication of slowing AI spending. In its July second-quarter earnings call, the company increased its 2025 expense guidance to between $114 billion and $118 billion. Meta also cautioned that its AI efforts will fuel expense growth in 2026 at a rate higher than this year’s gains.
The firm keeps investing in infrastructure projects, such as a historic $27 billion deal Tuesday with Blue Owl Capital to finance its Hyperion data center in rural Louisiana. Zuckerberg said the facility will be of such size that it will be able to span much of Manhattan’s footprint, giving an idea of the massive scale of Meta’s ambitions in AI.
The layoffs are part of Meta’s efforts to run more nimbly and double down on AI investment. Consolidating Wang’s power and removing redundancies, the firm aims to speed up its AI development and compete more effectively in a more competitive environment.
Meta is set to announce third-quarter earnings results next week, giving more insight into how these strategic changes are impacting the company’s financials and outlook. The restructuring indicates that although Meta continues to be committed to artificial intelligence, it’s willing to cut hard on personnel to make sure its investments bear fruit.
