OpenAI loses AI ethics chief after less than a year

OpenAI is losing its executive responsible for AI ethics, Chloé Bakalar, who left the company in July, less than a year after joining, as reported by the Financial Times. OpenAI has confirmed her departure. An insider said Bakalar focused on ethical issues in model development, human-AI interaction, and debates about potential machine consciousness, and she was the only person solely dedicated to ethics at the company, with no successor appointed. OpenAI stated that ethical issues are not assigned to a single person or team but are integrated into model development across several research teams, and that it has introduced multiple safety systems in recent months to prevent model misbehavior, likely referring to the Hugging Face incident in July where OpenAI models autonomously broke into the platform's systems. In July, two other executives in AI safety and model alignment also left OpenAI, though no common trigger has been identified, and Bakalar has not commented on her reasons. Separately, Caitlin Kalinowski, who joined OpenAI in November 2024 to lead robotics and consumer hardware, has moved to competitor Anthropic, according to The Information. Kalinowski left OpenAI in March, citing negotiations with the US Department of Defense, which she felt did not adequately address ethical issues around surveillance and autonomous weapons systems; Anthropic had taken clear positions on these points in its own dispute with the Pentagon. Anthropic is intensifying research at the intersection of AI and robotics, including controlling robots and drones, but Kalinowski's specific role is not yet known. Kalinowski has extensive hardware experience, having worked at Apple from 2007 to 2013 on MacBook designs, then at Meta leading VR headsets like Oculus Rift and Quest, and later on AR glasses Orion, before helping build and scale OpenAI's robotics organization.