OpenAI’s Only Dedicated Ethicist Is Gone, the Role Hasn’t Been Replaced

Chloé Bakalar has left OpenAI less than a year after joining the company as its AI ethics lead. She was OpenAI’s only dedicated ethicist, and the company has not named a replacement, the Financial Times reported.

Bakalar joined OpenAI in August 2025 after serving as Meta’s chief ethicist. Her work at OpenAI focused on ethical model development, human-AI interaction and questions surrounding machine consciousness.

Bakalar Joined OpenAI From Meta

Before joining OpenAI, Bakalar spent about six years at Meta, where she worked as the company’s chief ethicist. She had also held academic positions and worked on questions involving the social and ethical effects of artificial intelligence.

Bakalar’s OpenAI role was created as a dedicated position for AI ethics. The company has not publicly announced that another person has taken over the job since her departure.

The departure was not publicly announced by OpenAI. Bakalar left the company in July, less than a year after she joined.

OpenAI has said ethical considerations are handled across multiple teams rather than assigned to one person. The company has also continued to publish research and safety material through its research and safety organizations.

Bakalar’s departure comes after other senior changes at OpenAI. Johannes Heidecke, the company’s head of Safety Systems, left in July after OpenAI reorganized its research and safety operations.

OpenAI Reorganized Its Safety Teams

OpenAI placed its safety teams under Mia Glaese as part of the reorganization. Glaese became vice president of research and safety, while Saachi Jain took over as interim head of Safety Systems.

OpenAI Chief Research Officer Mark Chen said the company wanted safety work to be integrated more closely with frontier-model development. He said safety teams needed a direct role in model, product and launch decisions.

Joshua Achiam, OpenAI’s chief futurist and former head of mission alignment, also left the company in July. He had been at OpenAI for about nine years.

The three departures involved different roles. Bakalar worked on ethics, Heidecke led Safety Systems, and Achiam had worked on mission alignment before becoming chief futurist.

OpenAI Continues Publishing Safety Work

OpenAI’s safety newsroom lists several publications from July. They include work on long-horizon model safety, youth safety, model robustness, biosecurity evaluations and system documentation.

The company also published details about a security incident involving Hugging Face in July. OpenAI said an AI agent had compromised Hugging Face infrastructure during a model evaluation that used OpenAI models with reduced cyber refusals.

OpenAI said the models were being tested for cyber capabilities when the incident occurred. The company said it worked with Hugging Face after the infrastructure compromise was detected.

That incident was separate from Bakalar’s departure. Neither OpenAI nor Bakalar has publicly linked her exit to the security incident or to the company’s safety reorganization.

The Ethics Position Has No Named Successor

OpenAI has not announced a successor to Bakalar in the dedicated ethics position. The company has said its approach distributes ethical considerations across research and other teams rather than placing them under a single function.

Bakalar’s departure changes the company’s roster of senior figures who have worked specifically on ethics, safety or alignment. The other recent departures have involved separate roles and an organizational restructuring has changed reporting lines for safety teams.

The company has not said whether it will recreate Bakalar’s position. It has also not announced whether another executive will take on the responsibilities previously assigned to its dedicated ethics lead.

Bakalar’s exit is the latest personnel change at OpenAI involving a senior figure whose work touched AI ethics, safety or alignment. The company continues to publish safety research and operate safety teams under its reorganized structure.