Meta Taps Ex-OpenAI Scientist Shengjia Zhao to Lead Its Superintelligence AI Lab

Hans India

26-07-2025

In a bold move that signals Meta's intensified push into advanced artificial intelligence, CEO Mark Zuckerberg has announced the appointment of Shengjia Zhao—an ex-OpenAI researcher and co-author of the original ChatGPT paper—as Chief Scientist of Meta's Superintelligence AI group.
Zhao, who quietly joined the tech giant in June, played a pivotal role in shaping OpenAI's early breakthroughs, including the development of its first reasoning model known as o1. That model was instrumental in triggering the now widespread 'chain-of-thought' trend, later adopted by other AI powerhouses such as Google and DeepSeek.
Confirming Zhao's new leadership role, Zuckerberg wrote in a post on Threads: 'Our lead scientist from day one. Now that our recruiting is going well, and our team is coming together, we have decided to formalise his leadership role.'
Zhao will now report directly to Alexandr Wang, the former Scale AI CEO, who was appointed Meta's Chief AI Officer in June. Wang oversees Meta's high-stakes pursuit of artificial general intelligence (AGI)—intelligent systems capable of reasoning and problem-solving at or above human levels.
This appointment places Zhao at the helm of Meta's recently launched Superintelligence Lab, a standalone division separate from the long-established FAIR (Fundamental AI Research) unit led by AI veteran Yann LeCun. While LeCun continues to lead FAIR, he now reports to Wang, giving the new Chief AI Officer unified control over Meta's dual-track AI research strategy.
Meta's aggressive AI talent acquisition has been in full swing this summer. In a matter of weeks, the company has hired over a dozen researchers from rivals like OpenAI, Google, Apple, and Anthropic. Along with Zhao, several high-profile AI minds—including Jiahui Yu, Shuchao Bi, and Hongyu Ren—have transitioned from OpenAI to Meta. Other significant additions include Trapit Bansal, Zhao's past collaborator, and three multimodality experts from OpenAI's Zurich office.
According to reports, Zuckerberg himself has led the recruitment campaign, personally contacting candidates via email and even hosting them at his Lake Tahoe retreat. The stakes are high—many of the offers are said to include hefty eight- and nine-figure compensation packages, often with 'exploding' clauses that push for rapid acceptance.
Despite the buzz, Meta's current flagship AI model, LLaMA 4, still trails behind OpenAI's GPT-4 and Google's Gemini in terms of performance. However, all eyes are now on 'Behemoth,' Meta's code-named next-gen model expected to launch later this year.
Zuckerberg remains confident about the road ahead. 'Together we are building an elite, talent-dense team that has the resources and long-term focus to push the frontiers of superintelligence research,' he stated.
With Zhao now officially at the helm and a wave of top-tier AI talent joining the ranks, Meta is clearly positioning itself as a serious contender in the race for the next big breakthrough in artificial intelligence.