Renowned university professors and academic researchers keep joining frontier AI labs such as OpenAI and Anthropic. Remaining faculty say the brain drain is threatening the development of open source AI models in the West.

So far this year, at least 22 professors and researchers have taken leave or stepped back from their roles at Stanford, Berkeley, Harvard, Virginia and the University of Southern California to work at OpenAI, Anthropic, Google, and Meta Platforms, according to a review of college websites, X posts and LinkedIn profiles. The actual number of departures could be far higher, according to my conversations with professors at some of the universities.

Many of the professors are computer scientists, but some are physicists, economists, statisticians or mathematicians. Two of them taught philosophy, which probably has something to do with the philosophical questions that arise when labs shape the personalities and self-awareness of AI. (See the list below.)

Just last week, Jelani Nelson, the chair of Berkeley’s department of electrical engineering and computer science, and UT Austin philosophy professor Harvey Lederman said they’d be taking leave to join Anthropic as members of its technical staff. (Lederman will continue teaching at New York University.)

AI professors have been leaving school for big jobs at Google and its ilk since deep learning and self-driving software got hot 15 years ago, and after the generative AI boom got rolling a couple of years ago. But Berkeley computer science professor and AI researcher Joey Gonzalez said the exits have accelerated this year.

The reason is the same as it’s always been: It’s simply more difficult to do cutting edge research in academia because it requires large amounts of expensive AI servers, which the schools generally don’t have.