Meta CEO Mark Zuckerberg announced Friday that Shengjia Zhao, a former OpenAI researcher, will serve as Chief Scientist of Meta Superintelligence Labs (MSL), the company’s newly formed AI division. Zhao played a pivotal role in several of OpenAI’s most significant advancements, including ChatGPT, GPT-4, and the company’s first AI reasoning model, o1.
“I’m excited to share that Shengjia Zhao will be the Chief Scientist of Meta Superintelligence Labs,” Zuckerberg wrote in a post on Threads. “Shengjia co-founded the new lab and has been our lead scientist from the beginning. Now that our recruiting efforts are progressing and the team is coming together, we’ve decided to officially formalize his leadership.”
Zhao will lead MSL’s research agenda under the broader direction of Alexandr Wang, the former Scale AI CEO, who was recently appointed to head the new unit.
While Wang’s non-research background made his appointment somewhat unconventional, Zhao’s addition brings strong technical and academic expertise to the team. Zhao is widely respected for his work on cutting-edge AI models, and his presence helps round out MSL’s leadership. To strengthen the unit further, Meta has recruited top-tier talent from OpenAI, Google DeepMind, Anthropic, Safe Superintelligence, Apple, and its own FAIR (Fundamental AI Research) and generative AI divisions.
In his announcement, Zuckerberg highlighted Zhao’s groundbreaking contributions, including a “new scaling paradigm”—likely a reference to Zhao’s foundational work on OpenAI’s o1 reasoning model, co-developed with OpenAI co-founder Ilya Sutskever. As Meta currently lacks a direct competitor to o1, building advanced reasoning models has become a strategic focus for MSL.
According to a June report by The Information, Zhao joins MSL alongside three other prominent OpenAI researchers — Jiahui Yu, Shuchao Bi, and Hongyu Ren. Meta has also recruited Trapit Bansal, another former OpenAI researcher who worked with Zhao, and several researchers from OpenAI’s Zurich office specializing in multimodal AI.
Zuckerberg is going all-in on making MSL a top-tier research lab. Reports suggest he has personally reached out to AI researchers, even hosting some at his Lake Tahoe estate, and Meta is offering aggressive compensation packages — including eight- and nine-figure deals with limited windows to accept.
To support the lab’s massive computational needs, Meta is ramping up its investment in cloud infrastructure. By 2026, MSL will gain access to Prometheus, Meta’s 1-gigawatt AI cloud cluster in Ohio. Once operational, Prometheus will be among the largest AI training clusters globally, offering enough computing power to serve over 750,000 homes — and enabling MSL to run the large-scale training needed for frontier AI development.
With Zhao onboard, Meta now has two top AI scientists: Zhao at MSL and Yann LeCun, who leads FAIR. While MSL focuses on near-term, competitive AI systems, FAIR continues to work on long-term foundational research intended for breakthroughs five to ten years into the future. The interaction between Meta’s three AI-focused groups — MSL, FAIR, and the generative AI unit — remains to be clarified.
Still, with Zhao’s appointment, Meta has solidified a formidable leadership team, positioning itself as a serious contender in the race against OpenAI and Google in the next wave of AI innovation.



