Mark Zuckerberg, CEO of Meta, has established a new initiative called Meta Superintelligence Labs (MSL), according to reports. This formation appears to be part of a wider reorganization of the existing Meta AI division. The MSL team will focus on pioneering advancements in large language models (LLMs) and artificial intelligence (AI) technologies for the company. Alexandr Wang, former CEO of Scale AI, has been appointed to lead this new group, having joined Meta earlier this month.
A report from Bloomberg highlights that Zuckerberg communicated the restructuring and the inception of MSL to Meta employees via an internal memo. The document reportedly confirms Wang’s leadership position within the AI division, with Zuckerberg referring to him as the “most impressive founder of his generation.”
In a notable collaboration, Nat Friedman, the previous CEO of GitHub, which was acquired by Microsoft, will be closely partnering with Wang to steer MSL. The initiative has already made significant strides by recruiting 11 new team members, many of whom are coming from industry competitors including OpenAI, Anthropic, and Google DeepMind.
Wang took to X (formerly known as Twitter) to announce his new role as Chief AI Officer at Meta, alongside sharing the identities of the newly onboarded members of MSL. Notable hires include talents from OpenAI, such as Trapit Bansal, co-creator of the o-series models; Shuchao Bi, involved in developing GPT-4o’s Voice Mode; and Huiwen Chang, who contributed to the image generation capabilities of GPT-4o. Additionally, the team comprises former employees from DeepMind and Anthropic. Reports suggest that the compensation for these new recruits is in the eight-figure range. Below is the complete list of MSL team members:
| Name | Role | Company |
|---|---|---|
| Trapit Bansal | Pioneered RL on chain of thought; co-creator of o-series models | OpenAI |
| Shuchao Bi | Co-creator of GPT-4o Voice Mode and o4-mini; led multimodal post-training | OpenAI |
| Huiwen Chang | Co-creator of GPT-4o’s image generation; invented MaskGIT and Muse architectures | Google Research |
| Ji Lin | Helped build o3/o4-mini, GPT-4o, GPT-4.1, GPT-4.5, 4o-imagegen, Operator reasoning stack | OpenAI |
| Joel Pobar | Inference at Anthropic; ex-Meta on performance and ML tools | Anthropic |
| Jack Rae | Pre-training lead for Gemini; reasoning for Gemini 2.5; led Gopher and Chinchilla | DeepMind |
| Hongyu Ren | Co-creator of GPT-4o, 4o-mini, o1-mini, o3-mini, o3 and o4-mini; led post-training | OpenAI |
| Johan Schalkwyk | Former Google Fellow; early contributor to Sesame; lead for Maya | |
| Pei Sun | Post-training, coding, reasoning for Gemini; ex-Waymo perception model creator | Google DeepMind |
| Jiahui Yu | Co-creator of o3, o4-mini, GPT-4.1, GPT-4o; led perception team; co-led multimodal | OpenAI / Gemini |
| Shengjia Zhao | Co-creator of ChatGPT, GPT-4, all mini models, 4.1 and o3; led synthetic data | OpenAI |
In remarks to Bloomberg, Zuckerberg emphasized the urgency of AI advancements, stating, “As the pace of AI progress accelerates, developing superintelligence is coming into sight. I believe this will be the beginning of a new era for humanity, and I am fully committed to doing what it takes for Meta to lead the way.”
