Dossier
Dylan Scand
Coverage of Dylan Scand in the Nexus archive.
- The leaders responsible for keeping OpenAI's AI safe keep leaving
OpenAI's head of safety systems, Johannes Heidecke, has left the company, joining a growing list of safety team leaders who have departed. Departing researchers have criticized OpenAI's focus on product releases over safety, while the company reorganizes safety efforts under Mia Glaese. Anthropic, a rival AI firm, has also experienced safety team departures.