OpenAI experienced a significant talent exodus in 2025, with at least 11 key scientists and executives leaving the organization, many of whom joined Meta’s Superintelligence Lab. This shift occurred during a time of transformation for OpenAI, which had already seen the departure of several high-profile executives the previous year.
Among the notable departures in the summer of 2025 were more than half a dozen researchers who transitioned to Meta, which has invested heavily in expanding its AI capabilities. This move aligns with Meta’s goal to enhance its artificial intelligence initiatives, particularly at its Superintelligence Lab, which has attracted top talent from the industry.
Key Departures Impacting OpenAI
The summer months proved particularly challenging for OpenAI, as prominent figures like Jason Wei, Zhiqing Sun, and Hyung Won Chung made their way to Meta. Wei, who had been instrumental in developing OpenAI’s o1 and deep research models, left in July. Sun, also a research scientist at OpenAI, joined Meta’s Superintelligence Lab during the same period.
Chung, who was part of a trio of researchers departing for Meta, expressed enthusiasm about their new venture, stating on LinkedIn that they were “having so much fun building from a clean slate with a truly talent-dense team.” These sentiments reflect the excitement surrounding Meta’s ambitious AI projects.
Another significant loss for OpenAI was Shengjia Zhao, who became the chief scientist of the Meta Superintelligence Lab after co-creating ChatGPT and GPT-4. Zhao is now collaborating directly with Mark Zuckerberg and Alexandr Wang, Meta’s Chief AI Officer.
Furthermore, Jiahui Yu, a leader in developing the perception capabilities of large language models, left for Meta in late June. His expertise in integrating images, audio, and sensor readings into AI models will likely bolster Meta’s ongoing efforts in the field.
Other researchers who transitioned to Meta include Hongyu Ren, who was integral to the development of OpenAI’s GPT-4o model, and Shuchao Bi, who worked on multimodal and reinforcement learning.
Executive Departures and Future Implications
The talent drain did not stop with researchers; OpenAI also saw the exit of several executives. In November, Larry Summers, a former Treasury Secretary and Harvard president, resigned from the OpenAI board following the release of emails linked to Jeffrey Epstein. This resignation added to the company’s challenges during a restructuring phase.
In August, Julia Villagra stepped down from her role as chief people officer, having been promoted to the position only months earlier in March. Additionally, Liam Fedus, who served as vice president of research and post-training, left in March to co-found an AI startup called Periodic Labs, focusing on creating an AI scientist.
OpenAI’s data scientist, Tom Cunningham, also resigned in November, joining a non-profit research institute dedicated to evaluating the safety and capabilities of AI models.
Finally, as 2025 drew to a close, Hannah Wong, OpenAI’s chief communications officer, announced her departure, seeking her “next chapter.” She indicated that Lindsey Held Bolton would lead the communications team on an interim basis while the company searches for a new chief communications officer.
As OpenAI navigates these high-profile exits, CEO Sam Altman remains one of only two active members from the company’s original founding team of 11. The ongoing departures raise questions about the future direction of OpenAI and its ability to retain talent in an increasingly competitive AI landscape.
