Meta CEO Mark Zuckerberg has confirmed that Shengjia Zhao, a former OpenAI researcher, will now officially head research efforts at Meta Superintelligence Labs (MSL). Zhao, a key contributor to OpenAI breakthroughs such as ChatGPT, GPT-4, and the AI reasoning model, o1, has been with MSL since its inception.
In a Threads post on Friday, Zuckerberg announced, “I’m thrilled to share that Shengjia Zhao will serve as the Chief Scientist of Meta Superintelligence Labs. Shengjia helped co-found this new division and has led our scientific direction from the beginning. Now that our hiring efforts are gaining momentum and the team is taking shape, we’re formalizing his leadership.”
We are excited to announce that @shengjia_zhao will be the Chief Scientist of Meta Superintelligence Labs!
— Alexandr Wang (@alexandr_wang) July 25, 2025
Shengjia is a brilliant scientist who most recently pioneered a new scaling paradigm in his research. He will lead our scientific direction for our team.
Let's go 🚀 pic.twitter.com/D93KQWIvFl
Under the broader leadership of Alexandr Wang, the former CEO of Scale AI, Zhao will define the research path for MSL. Although Wang comes without a formal research background, his strategic capabilities are complemented by Zhao’s deep technical expertise. Together, they are expected to drive the lab’s mission forward.
To strengthen MSL further, Meta has onboarded top-tier AI talent from organizations like OpenAI, Google DeepMind, Anthropic, Apple, and Safe Superintelligence. It has also transitioned researchers from its in-house Fundamental AI Research (FAIR) lab and generative AI teams to MSL.
Zuckerberg emphasized Zhao’s role in pioneering innovations, notably referencing a “new scaling paradigm,” which likely points to Zhao’s influence on OpenAI’s AI reasoning model, o1. Since Meta doesn’t currently have an equivalent to o1, this area has become a focal point for Meta Superintelligence Labs.
In June, The Information disclosed that Zhao would be joined at Meta Superintelligence Labs by fellow OpenAI researchers Jiahui Yu, Shuchao Bi, and Hongyu Ren. The company has also brought in Trapit Bansal, another contributor to OpenAI’s reasoning models, along with three researchers from OpenAI's Zurich team focused on multimodal systems.
Zuckerberg is reportedly taking a hands-on approach to ensure MSL’s success. He’s actively reaching out to top researchers — even inviting them to his Lake Tahoe residence — and offering highly competitive packages. Some of these offers reach eight or nine figures, with “exploding” clauses that require quick decisions.
Meta is also scaling its infrastructure investments to support the lab's mission. This includes enhancing its cloud capabilities to accommodate the large-scale training required for advanced AI development.
By 2026, Zhao and the team at MSL will gain access to Meta’s 1 gigawatt cloud computing cluster, Prometheus, situated in Ohio. Once operational, Prometheus will place Meta among a rare few tech giants capable of running such intensive AI training workloads. To put its power into context — 1 gigawatt can supply electricity to more than 750,000 homes.
With Shengjia Zhao joining Yann LeCun, head of Meta’s FAIR lab, Meta now has two top AI scientists in its leadership structure. While FAIR continues to explore long-term AI innovation — research aimed 5 to 10 years into the future — the operational balance and collaboration between MSL, FAIR, and Meta’s generative AI group remains to be fully defined.
Still, Meta appears poised with a strong, competitive AI leadership team — ready to challenge the likes of OpenAI and Google in the race for frontier AI dominance.