Zuckerberg announces Meta Superintelligence Labs

Mark Zuckerberg told employees in an internal memo that Meta was consolidating its AI efforts — the open-weight Llama model line, the Fundamental AI Research (FAIR) group, and the teams building AI into Meta’s products — into a single new division called Meta Superintelligence Labs. Alexandr Wang, who had joined Meta weeks earlier as part of its $14.3 billion investment in Scale AI, took the newly created role of Chief AI Officer to lead the group; Zuckerberg said former GitHub chief executive Nat Friedman would partner with Wang to run it.

The memo named eleven researchers Meta had newly hired to join the effort, drawn from rival labs including OpenAI, Google DeepMind and Anthropic — part of a hiring push that had drawn public attention, and some criticism from competitors, for offering unusually large compensation packages to individual researchers. Zuckerberg called Wang “the most impressive founder of his generation” and wrote that “as the pace of AI progress accelerates, developing superintelligence is coming into sight,” framing the reorganisation as a response to Meta’s AI division lagging what he saw as faster progress at OpenAI and Google.

The restructuring followed a period in which Meta’s Llama models had drawn criticism for falling behind competitors on benchmarks and in developer adoption despite the company’s open-weight strategy, and came after earlier, less drastic reorganisations of its AI teams had not resolved that gap. Combined with the Scale AI investment three weeks earlier, the moves represented one of the most direct examples in 2025 of an established technology company restructuring around an acquired leader and imported talent rather than promoting from within, a pattern echoed elsewhere in the industry that year.