Meta CEO Mark Zuckerberg has recently revealed a significant development in the company’s ambitious plans for artificial intelligence. Shengjia Zhao, a respected figure in the AI community and a former researcher at OpenAI, has been officially appointed as the Chief Scientist of Meta's newly established Superintelligence AI group. Zhao played a major role in OpenAI’s early breakthroughs, particularly in developing the original reasoning model known as o1, which set the foundation for the popular chain-of-thought prompting techniques now widely used in the AI field. He also contributed to shaping ChatGPT, making his move to Meta a strategic victory for the tech giant. Although Zhao quietly joined Meta in June, his leadership role was only now formally announced by Zuckerberg through a post on Threads, where he praised Zhao as the lead scientist of their superintelligence efforts from the very beginning.
This appointment is part of Meta’s larger, aggressive recruitment campaign aimed at establishing itself as a dominant player in the evolving AI landscape. Zhao will now report directly to Alexandr Wang, formerly the CEO of Scale AI, who was brought on board in June as Meta’s Chief AI Officer. Wang’s mission is to guide the company toward the development of artificial general intelligence (AGI)—a highly advanced form of AI capable of thinking, reasoning, and solving problems like, or even beyond, human beings. Zhao’s elevation comes as Meta’s Superintelligence Lab, launched just a month earlier in June 2025, gains momentum. This new lab functions independently from Meta’s older and well-established AI division, FAIR (Facebook AI Research), which is still led by AI luminary Yann LeCun. Notably, LeCun now reports to Wang, placing Meta’s AI divisions under a unified leadership structure to accelerate progress toward AGI.
Meta’s intense drive to poach top AI talent has been one of the most talked-about stories in the tech world throughout the summer. The company has managed to hire more than a dozen prominent researchers from competitors like OpenAI, Google, Apple, and Anthropic. Notable hires include Apple AI experts Tom Gunter and Mark Lee, along with three other OpenAI scientists—Jiahui Yu, Shuchao Bi, and Hongyu Ren—who joined Meta around the same time as Zhao. Another significant recruit is Trapit Bansal, who had previously collaborated with Zhao on AI reasoning models. Additionally, Meta welcomed a trio of engineers from OpenAI’s Zurich office, all with specialized knowledge in multimodal systems that can handle text, images, and video inputs together.
What makes this hiring spree even more extraordinary is the personal involvement of Zuckerberg himself. He has reportedly reached out to prospective candidates directly, sometimes even inviting them to his private Lake Tahoe retreat for in-person discussions. Some of the offers extended by Meta are said to be remarkably generous, including compensation packages worth eight to nine figures. These offers often come with high-pressure deadlines, known as “exploding offers,” which force candidates to make quick decisions—often within just a few days. This hands-on, high-stakes approach underscores just how serious Meta is about leading the next wave of AI development.
Currently, Meta’s most sophisticated open-source AI model, LLaMA 4, is still behind competitors such as OpenAI’s GPT-4 and Google’s Gemini. However, the company is placing heavy bets on its upcoming model, codenamed “Behemoth,” which is expected to launch later this year. Mark Zuckerberg remains confident in Meta’s direction, emphasizing that the company is assembling a top-tier team with the resources, expertise, and vision necessary to drive innovation in superintelligent systems. With Shengjia Zhao now at the helm of Meta’s most advanced AI research group, and a steadily growing roster of elite scientists and engineers joining the ranks, the company is clearly positioning itself to achieve the next groundbreaking leap in artificial intelligence—and to ensure that leap carries Meta’s brand.
