Meta will lay off around 600 employees from its artificial intelligence (AI) division as part of an organisational restructuring aimed at streamlining operations and reducing management layers, according to a report by Reuters.
The decision was outlined in an internal memo from Alexandr Wang, Meta’s Chief AI Officer, who joined the company in June 2025 following Meta’s $14.3 billion investment in Scale AI. The layoffs will affect employees across several divisions, including AI infrastructure, the Fundamental Artificial Intelligence Research (FAIR) unit, and other AI-related product teams.
Selective retention of key hires
Employees in TBD Labs, which includes many of Meta’s recent high-profile AI recruits working on large language models, will not be impacted. These hires, brought in under Wang’s leadership earlier this year, are considered central to Meta’s long-term AI roadmap.
Insiders said the company’s AI division had become “overextended,” with FAIR and product-oriented teams competing for limited computing resources. The newly created Superintelligence Labs, established under Wang’s supervision, has inherited much of this structure. The layoffs are intended to reduce duplication of work and consolidate control of Meta’s AI direction under Wang’s leadership.
Realigning focus on AI advancement
Meta continues to expand its AI initiatives to compete with OpenAI and Google. The company has invested heavily in AI infrastructure and recruitment, with Superintelligence Labs now employing just under 3,000 people after the workforce reduction.
Affected staff were informed that their official termination date will be 21 November 2025 and that they are currently in a non-working notice period. During this time, employees will retain pay and benefits but have restricted internal access. Meta will provide 16 weeks of severance pay plus two additional weeks per completed year of service, adjusted for the notice period.
Leadership changes and continued investment
The restructuring follows internal concerns about the company’s AI progress, particularly after the Llama 4 model release in April drew limited response from developers. In response, CEO Mark Zuckerberg launched Meta Superintelligence Labs, led jointly by Wang and former GitHub CEO Nat Friedman, to accelerate AI development.
In July, Meta projected 2025 expenses between $114 billion and $118 billion, citing continued spending on AI initiatives. Earlier this week, the company announced a $27 billion partnership with Blue Owl Capital to fund the Hyperion data centre in Louisiana, which Zuckerberg said would cover a “significant part of Manhattan’s footprint.”
