Meta reorganizes AI division into Meta Superintelligence Labs, with new focus on foundation models and internal restructuring

Meta has officially restructured its AI organization, unveiling a new framework designed to accelerate its pursuit of superintelligence. The changes, confirmed through an internal memo and reported by Bloomberg and The New York Times, follow a series of leadership moves and strategic shifts aimed at closing the gap with competitors like OpenAI, Anthropic, and Google DeepMind. The overhaul centers on the creation of Meta Superintelligence Labs, or MSL, a new umbrella organization led by Alexandr Wang, who joined Meta in June as chief AI officer. Wang, the founder of Scale AI, brings deep expertise in AI data infrastructure and model training—key areas where Meta has sought to strengthen its capabilities. Within MSL, the most prominent new unit is TBD Labs, headed by Wang himself. This group will focus on foundational AI models, including the Llama series, which recently saw its latest release in April. The name “TBD” reflects the exploratory and forward-looking nature of the work, emphasizing innovation in core AI systems. The remaining three groups under MSL are dedicated to research, product integration, and infrastructure. These units are intended to streamline collaboration across development, deployment, and scaling of AI technologies. By separating these functions, Meta aims to improve agility and coordination, particularly as it pushes toward more advanced AI systems. This restructuring marks the latest in a series of organizational shifts at Meta as it seeks to regain momentum in the AI race. In recent months, the company has ramped up hiring, with CEO Mark Zuckerberg personally involved in recruiting top talent, according to a June report from Bloomberg. The move comes amid growing pressure after Meta’s public AI model releases have lagged behind those of its rivals. With the new structure in place, Meta is signaling a renewed focus on long-term AI advancement, positioning itself not just to compete but to lead in the next phase of artificial intelligence development.