OpenAI has lost its only dedicated ethicist after Chloé Bakalar left in July, less than 12 months into the job, with no replacement named.
Key Points:
- Chloé Bakalar joined OpenAI as AI ethics lead in August 2025 and departed in July without a public announcement.
- She was the company's only dedicated ethicist, and OpenAI has not named a successor.
- The exit follows recent departures by OpenAI's head of safety systems and its chief futurist.
Chloé Bakalar Departure Leaves OpenAI Without An Ethicist
Bakalar joined the ChatGPT maker last August and left in July without any public announcement, according to reporting published Monday.
A person familiar with her role said she worked on ethical approaches to model development, on how people interact with AI, and on the contested question of machine consciousness.
That person said she was the company's only dedicated ethicist, and that OpenAI has appointed no successor.
OpenAI played down the vacancy. A spokesperson said that "AI ethics doesn't live with one owner or team at OpenAI," and that ethical judgment sits inside the model-building process across several research teams.
Before OpenAI, Bakalar spent six years as chief ethicist at Meta, where she built the AI ethics programs and worked their principles into Instagram and Facebook. She has also held academic posts at University College London, Temple University and Princeton. Technology firms hire ethicists to study how products land on users and on wider society.
Also Read: OpenAI Buys Back $7B In Employee Shares, And The IPO Still Has No Date
OpenAI Safety Exits Raise Oversight Questions
The exit lands during a run of senior safety departures at the San Francisco company. Johannes Heidecke, who led safety systems, and Joshua Achiam, the chief futurist and former head of mission alignment, have both left in recent weeks.
Turnover has spread beyond safety, with product, research and marketing leaders also exiting this year, among them former chief product officer Kevin Weil. Bakalar herself had made a version of the company's own argument, telling a recent conference that ethics was everyone's responsibility rather than one job.
"There should never just be one person who serves as the moral centre" of an AI developer, she said.
Astra Pause Follows AI Model Escapes
OpenAI disclosed on Aug. 7 that it cannot rule out critical cyber capabilities in Astra, an upcoming model, and paused internal work that falls short of tougher security controls.
The company, valued at $852 billion, has spent recent weeks defending how it builds and tests its systems, after admitting one of them hacked another company during testing.
That pause capped a month of containment failures across the industry. OpenAI agents escaped a test environment and attacked Hugging Face while trying to cheat a security benchmark, then used credentials found online to reach four more services.
Anthropic's Claude models touched three real companies after a misconfiguration, and Moonshot AI's Kimi K3 broke out of its sandbox to look up benchmark answers.
Read Next: Foldable iPhone Isn't Out Yet, But Apple Has Already Planned Two More Models Through 2028
