Shengjia Zhao, one of the key figures behind OpenAI’s major advancements, including GPT-4 and the o1 reasoning model, has been named Chief Scientist of Meta Superintelligence Labs (MSL), CEO Mark Zuckerberg confirmed on Friday.

Zuckerberg made the announcement in a post on Threads, noting that Zhao has led the lab’s research since its founding and will now formally step into a leadership position. Zhao will work under Alexandr Wang, the former Scale AI CEO who was tapped to lead MSL earlier this year.

Zhao’s appointment highlights Meta’s growing focus on AI reasoning, especially as the company develops a rival to OpenAI’s o1 model. Meta’s MSL division is rapidly recruiting top AI researchers, including four other former OpenAI scientists: Jiahui Yu, Shuchao Bi, Hongyu Ren, and Trapit Bansal.

By 2026, Zhao and the MSL team will gain access to Prometheus, Meta’s massive 1-gigawatt training cluster based in Ohio, capable of supporting advanced frontier model training.

Zhao now joins Yann LeCun as one of Meta’s two top AI scientists. LeCun continues to lead Meta’s FAIR division, which focuses on longer-term AI technologies. Coordination between FAIR and MSL is still evolving.