Sebastien Bubeck, a key figure in Microsoft's AI research, has confirmed his departure from the tech giant to join OpenAI. At Microsoft, Bubeck held the role of Vice President of Artificial Intelligence and Distinguished Scientist, playing a pivotal role in developing Microsoft's Phi series of compact language models.

During his tenure at Microsoft, Bubeck focused on generative AI, particularly on creating models designed to serve as more efficient alternatives to OpenAI's GPT models. His research led to the development of the smaller, cost-effective Phi models, some of which have been integrated into Microsoft's Bing chatbot and Office 365 Copilot, replacing OpenAI's models in certain applications.

It is currently unknown what specific role Bubeck will assume at OpenAI. In a statement, Microsoft mentioned that Bubeck "has decided to leave Microsoft to further advance" the development of general artificial intelligence. This move comes amid a series of personnel changes within OpenAI's research team, including the departure of former CTO Mira Murati last Friday.

Microsoft does not appear to be surprised by Bubeck's departure and has expressed intentions to maintain a collaborative relationship with him through OpenAI. Microsoft has made substantial investments in OpenAI. Meanwhile, Bubeck's former team at Microsoft plans to continue advancing the Phi models, which were previously described by Microsoft CEO Satya Nadella as a crucial component of Microsoft's AI strategy.

Bubeck's departure highlights the ongoing talent movement within the AI sector, where top experts frequently shift between leading tech companies or establish their own ventures. As the AI field continues to evolve, more such high-level changes that impact the industry's future are expected.