OpenAI Safety Executive Resigns Over Shift in Company Priorities
OpenAI safety lead Jan Leike steps down, citing a lack of resources for AI alignment research and a company culture that favors shiny products over safety. His departure follows the recent exit of co-founder and chief scientist Ilya Sutskever.
OpenAI safety executive Jan Leike resigns from his role leading the company's "superalignment" team after reaching a breaking point over leadership's core priorities. He states that building smarter-than-human machines is an inherently dangerous endeavor, but OpenAI's safety culture and processes take a backseat to shiny products.
Leike explains on X that his team sails against the wind and struggles to secure the computing power needed to conduct this crucial research. He emphasizes that over the past few months, the superalignment team is severely under-resourced, making it increasingly difficult to ensure advanced AI systems operate safely within human needs.
This high-profile exit comes amid a broader leadership shuffle at the startup, following the recent departure of OpenAI co-founder and chief scientist Ilya Sutskever. OpenAI CEO Sam Altman responds to the news by expressing appreciation for Leike's contributions, acknowledging that the company has a lot more to do regarding safety, and promising a longer post on the matter in the coming days.