Meta is undergoing another major restructuring of its AI organization, prompted by a sweeping internal memo from CEO Alexandr Wang, in which he warns that "superintelligence is coming." The memo, circulated across the company, outlines a dramatic shift in priorities, emphasizing the need for accelerated AI development and tighter integration of artificial intelligence across Meta’s core products. Wang frames the reorganization as a response to the rapidly evolving AI landscape, stressing that Meta must act decisively to remain competitive. The changes include consolidating AI teams, reallocating resources toward foundational model research, and empowering cross-functional units to fast-track innovation. While the move signals Meta’s commitment to leading in AI, it also reflects growing internal pressure to deliver on long-term promises amid rising competition from rivals like OpenAI and Google. The restructuring is expected to impact thousands of employees and could reshape the company’s technological trajectory for years to come.

Meta has unveiled a sweeping reorganization of its artificial intelligence operations, creating a new unit called Meta Superintelligence Labs (MSL), led by 28-year-old Alexandr Wang, the company’s newly appointed chief AI officer. The move, detailed in an internal memo, marks the most significant structural shift in Meta’s AI efforts to date, as the company accelerates its push toward achieving “superintelligence”—a point where AI surpasses human capabilities across nearly all intellectual domains. Wang’s memo outlines a new four-team structure: TBD Lab, FAIR (Facebook AI Research), Products & Applied Research, and MSL Infrastructure. TBD Lab, a small, elite team focused on training and scaling massive AI models, will be central to Meta’s long-term ambitions. It will explore “new directions,” including a mysterious “omni” model—likely a multimodal AI capable of processing text, audio, video, and other inputs simultaneously, aligning with MSL’s recent hires in diverse media domains. Shengjia Zhao, co-creator of ChatGPT, will serve as MSL’s chief scientist and lead research, though he is the only leader not reporting directly to Wang. FAIR, Meta’s longstanding research arm, will now function as an “innovation engine,” with its work directly feeding into TBD Lab’s model training. Rob Fergus will continue leading FAIR, while Yann LeCun, its chief scientist, will report to Wang—confirming earlier confusion about leadership structure. Nat Friedman, former GitHub CEO and investor, will lead product integration, guiding AI features into Meta’s apps and hardware like Quest VR headsets and AI glasses. Despite being initially named co-lead of MSL, Friedman now reports directly to Wang, signaling a consolidation of power under him. Aparna Ramani, a veteran Meta engineering executive, will head the newly unified MSL Infrastructure team, responsible for building the high-performance computing systems needed to train advanced models—critical given the massive GPU clusters required. This team absorbs infrastructure roles from both MSL and Meta’s broader AI operations. The reorganization also dissolves the AGI Foundations team, created in May to focus on long-term AI progress. Its members are being redistributed to FAIR, product teams, and infrastructure—though not to TBD Lab, raising questions about integration. This restructuring follows the dissolution of Meta’s GenAI division, which oversaw the Llama series, after the lukewarm reception of Llama 4. The constant churn reflects Meta’s urgency to catch up with rivals like OpenAI, Google DeepMind, and Anthropic, all of which have made rapid strides in AI development. Tensions have emerged within MSL, with some long-time researchers reportedly upset over the influx of high-paid new hires and perceived favoritism. Some have threatened to quit, according to prior reports. Wang emphasized that while organizational changes are disruptive, they are necessary for speed and focus. He stressed the importance of aligning research, infrastructure, and product development under a unified vision. The move underscores Meta’s aggressive strategy to dominate the next phase of AI, leveraging its vast resources, talent, and infrastructure. With Wang at the helm and a clear mandate to build superintelligence, Meta is betting that a bold, centralized approach will finally close the gap with its competitors—though the success of this reorg remains to be seen.