Why Apple’s AI Team Is Losing Top Talent to Meta, OpenAI, and Anthropic

Apple has recently lost four AI researchers, three of whom are of Chinese descent. Bloomberg’s Gurman reports that Jian Zhang, the chief AI researcher of Apple’s robotics group, is moving to Meta, a transition confirmed by Meta.

In addition to Zhang, three members of Apple’s foundation model team are slated to leave. The team’s morale appears low, with a reported loss of ten members in just a few weeks.

Jian Zhang, who headed the robot research unit within the AIML team, contributed a full stack of technologies from perception‑motion to emotional expression. He earned his bachelor’s at Zhejiang University and PhD at Purdue, spent a year teaching, and spent a decade at Apple before joining Meta as chief director of robotics at Meta Reality Labs.

Other departing staff include John Peebles, a core member of the foundation model team responsible for large‑language‑model training infrastructure and co‑author of the AXLearn paper; Nan Du, who led 3B‑scale edge model KV‑cache sharing, 2‑bit quantization, and PT‑MoE parallelism, and recently updated her LinkedIn to show a move to OpenAI; and Zhao Meng, who focused on training and optimizing large language models for Apple Intelligence and is heading to Anthropic after a background at Google Research.

The exodus follows a series of setbacks for Apple’s AI projects: the high‑profile Apple Intelligence demo at WWDC 2024, delays and bugs in a revamped Siri AI, and the admission in June 2025 that the Siri AI upgrade would be postponed to spring 2026. Internal reports suggest the turnover will continue.

Reference links: Bloomberg article 1 , Bloomberg article 2 .