Meta has recruited 4 researchers from OpenAI, adding new momentum to its superintelligence team and intensifying the competition among leading AI companies. The departures involve Shengjia Zhao, Shuchao Bi, Jiahui Yu, and Hongyu Ren, according to sources cited in the source article.
What Changed
Shengjia Zhao, Shuchao Bi, Jiahui Yu, and Hongyu Ren have joined Meta’s superintelligence team. Their OpenAI Slack profiles have been deactivated, a detail that underlines the finality of the move inside the company.
The Information first reported the departures. The source article says two sources confirmed to WIRED that the four researchers are leaving OpenAI for Meta.
The timing matters because Meta is in a broader push to close the gap with OpenAI, Anthropic and Google in the race to build artificial general intelligence. The latest hires are not described as routine staffing changes. Multiple sources at OpenAI told WIRED the departures were a major loss because the group was considered a key part of the company’s research strategy.
Why These Researchers Matter
The four researchers were connected to some of OpenAI’s most visible model work. Hongyu Ren was OpenAI’s post-training lead for the o3 and o4 mini models, according to sources cited in the article. Ren was also linked to the open source model that is set to be released this summer.
Post-training is the process of refining a model after it has already been trained on a primary dataset. That makes the role important because it sits close to how a model is shaped after its main training phase, including how it behaves and performs before release.
Jiahui Yu joined OpenAI in late 2023 after a stint at Google DeepMind. In Mark Zuckerberg’s memo, Yu is described as the “cocreator of o3, o4-mini, GPT-4.1 and GPT-4o.” The memo also says Yu previously led OpenAI’s perception team and formerly colead “multimodal at Gemini.”
Shuchao Bi was a manager of OpenAI’s multimodal models. Zuckerberg’s memo describes Bi as the “cocreator of GPT-4o voice mode and o4-mini.” Those details place Bi’s work in the area where AI systems handle more than one kind of input or output, including voice and multimodal model capabilities.
Meta’s Hiring Push
The departures follow comments from OpenAI CEO Sam Altman, who said earlier this month on a podcast that Meta had been making “giant offers” to OpenAI staffers. He described offers with “$100 million signing bonuses, more than that comp per year.”
Altman also said that “none of our best people have decided to take them up on that.” The latest moves now put that statement under sharper scrutiny, because multiple sources at OpenAI described the newly departed group as important to research strategy.
On Monday, Zuckerberg confirmed the moves in a memo to staff introducing the superintelligence team. The memo is part of a larger signal that Meta’s Superintelligence Lab is beginning to take form after a period of aggressive recruiting.
The source article also notes that OpenAI had recently lost three researchers from its Zurich office to Meta, according to The Wall Street Journal. Taken together, the latest departures suggest Meta is not only hiring broadly, but is also targeting people with direct experience on frontier AI models.
How OpenAI Is Responding
Inside OpenAI, the reaction appears serious. Mark Chen, the company’s chief research officer, wrote in a memo over the weekend that he felt as if someone had “broken into our home and stolen something.” He also told staff, “Please trust that we haven’t been sitting idly by.”
Those remarks show that OpenAI views the departures as more than a public relations setback. The issue is strategic: researchers with experience on o3, o4 mini, GPT-4.1, GPT-4o, GPT-4o voice mode, perception, post-training and multimodal systems are now moving to a direct competitor.
For OpenAI, the immediate challenge is retaining key talent while continuing its model roadmap. For Meta, the hires add technical depth to a superintelligence team that is being introduced with high-level leadership and direct attention from Zuckerberg.
What Comes Next
Meta’s Superintelligence Lab is now being organized under Alexandr Wang, whom Zuckerberg announced as the company’s “chief AI officer” and leader of the lab. Former GitHub CEO Nat Friedman is also part of the leadership structure described in the memo.
OpenAI and Meta did not immediately respond to a request for comment, according to the source article. That leaves the public picture largely defined by internal memos, source confirmations and the reported movement of researchers between the companies.
The core takeaway is simple: AI talent is becoming a decisive battleground. Meta has gained researchers tied to major OpenAI systems, while OpenAI is confronting the loss of people its own sources considered central to research strategy. In the race toward artificial general intelligence, the movement of a small number of researchers can carry outsized strategic meaning.