Meta Cuts About 600 AI Roles, Automates Compliance

Meta cut about 600 roles across its AI research, infrastructure and product teams—principally within FAIR and Superintelligence Labs—as part of a reorganization led by Chief AI Officer Alexandr Wang to make teams leaner and speed decision‑making; the new TBD Lab was reportedly spared. The reductions included senior researchers such as Tian Yuandong, who criticized the move and argued that those responsible for setbacks like Llama 4 were not dismissed. Meta also automated parts of its risk and privacy review, eliminating more than 100 compliance roles, which has raised internal concern that automated systems may miss nuanced regulatory and safety issues. Impacted staff were offered severance (reported as at least 16 weeks plus pay per year of service) and encouraged to seek internal roles, while startups such as Smallest AI publicly recruited displaced talent. The cuts follow a year of heavy AI hiring, large deals including a reported $14.3 billion Scale AI arrangement and multimillion‑dollar hire packages, and come as Meta refocuses on next‑generation models under a product‑focused vision shared by Wang and CEO Mark Zuckerberg.