Meta Platforms Inc. is eliminating approximately 600 positions in its artificial intelligence (AI) division as part of a reorganization aimed at streamlining operations and decision-making.
The cuts, as reported by Axios and the Wall Street Journal, will impact several teams across Meta’s AI unit, including the Facebook Artificial Intelligence Research (FAIR) group, product-focused AI teams, and AI infrastructure divisions. However, the company’s recently formed TBD Lab, a smaller unit working on next-generation foundation models that has attracted high-profile hires personally recruited by Meta CEO Mark Zuckerberg, will not be impacted.
The layoffs represent a fraction of the several thousand employees in Meta’s AI division. The company has encouraged impacted staff to apply for internal positions, framing the move as a strategic realignment rather than a wholesale reduction in workforce. In a memo to staff, Meta Chief AI Officer Alexandr Wang said the reduction would enable faster decision-making and give remaining employees greater responsibility and impact.
“Meta’s job-cutting moves are absolutely in response to the land rush of AI hires over the past few years,” said Mitch Ashley, vice president/practice lead of DevOps and AppDev at The Futurum Group. “It’s also an indication that the direction of the Superintelligence Labs is gaining some clarity. It too is a sign the the hiring land rush is slowing and Meta must focus on the competitiveness and revenues in a market that’s looking for tangible results from AI.”
The reorganization follows Meta’s establishment of Superintelligence Labs in June 2025, a move that came after the departure of senior staff and tepid response to its open-source Llama 4 model. The new structure consolidates FAIR, product-oriented AI teams, and TBD Lab under one umbrella, with the goal of focusing on long-term AI research, infrastructure development, and product innovation.
Meta has been building its AI capabilities for more than a decade, beginning with the creation of FAIR in 2013 under Chief AI Scientist Yann LeCun. The company has since expanded to a global research network focused on deep learning and has recruited top talent from competitors including OpenAI, Anthropic, and Google DeepMind, with some compensation packages reaching $100 million.
Meta’s AI strategy is backed by massive infrastructure investments. Last month, Meta and its partner Blue Owl Capital finalized a $27 billion financing agreement to build the Hyperion data center in Louisiana. The facility is projected to eventually require five gigawatts of power and reach an initial capacity of two gigawatts by 2030.
Zuckerberg has signaled Meta could invest more than $600 billion in the United States through 2028, depending on progress in AI development, underscoring the company’s long-term commitment to the technology.
“Meta’s move is just a look at what’s coming for every enterprise experimenting with autonomous decision-making,” Debra Andrews, CEO of consultancy Marketri, said in an email. “And the concern now isn’t just whether or not companies will be transparent, but whether they will be auditing results for bias and maintaining the human checkpoints that keep the tools aligned with ethical outputs.”
