.png){kind=logo}
OpenAI’s only dedicated AI ethicist, Chloé Bakalar, has left the company less than a year after joining. She joined as AI ethics lead in August 2025 and departed in July 2026. OpenAI did not announce the departure publicly, and no one has replaced her in the role.
Her exit adds to several recent changes across OpenAI’s safety and alignment leadership. Johannes Heidecke, who led safety systems, and Joshua Achiam, chief futurist and former mission alignment head, also left recently. The departures come as the company reviews safeguards for more capable artificial intelligence models.
Bakalar focused on ethical approaches to model development and the relationship between people and AI systems. Her work also covered debate around machine consciousness, according to a person familiar with her responsibilities. She served as the company’s only employee dedicated solely to ethics.
OpenAI thanked Bakalar for her work and said ethics responsibilities extend across the company. “We’re thankful for Chloé’s contributions,” a spokesperson said. The spokesperson added, “AI ethics doesn’t live with one owner or team at OpenAI.” Several research teams consider ethical questions during model development, the company said.
Before OpenAI, Bakalar spent six years at Meta and served as its chief ethicist. She helped create AI ethics programs and brought their principles into products, including Facebook and Instagram. Her work connected research with product design and policy questions affecting users.
Bakalar also held academic posts at University College London, Temple University and Princeton University. Technology companies hire ethicists to examine how products affect users and wider society.
Their work supports policy and regulatory discussions. That broad remit often connects early model design with later product decisions across several internal teams. It also informs management decisions at large developers.
At a recent AI conference, she described shared responsibility inside technology companies. “Ethics is actually everyone’s responsibility,” she told the audience. Bakalar declined to comment on her OpenAI departure.
Meanwhile, OpenAI faces closer examination of its safety controls as its models gain stronger cybersecurity abilities. The company recently confirmed that one model exploited a Hugging Face system during authorized testing. OpenAI clarified that Astra, an upcoming model, did not take part in that test.
OpenAI said recent Astra evaluations showed rapid gains in automated coding and cybersecurity tasks. The company could not rule out critical cyber capabilities under its Preparedness Framework. It paused Astra work that did not meet stronger security requirements and expanded safeguard testing.
Additionally, OpenAI introduced tighter access rules, isolated testing environments and broader monitoring for risky actions. The company also plans to work with government agencies and selected safety organizations during testing. These measures apply while engineers continue controlled development of Astra.
The company says its safety structure does not depend on a single ethics position. Its Preparedness Framework assigns internal safety leaders to review capability reports and safeguards. OpenAI leadership makes final model decisions after receiving those recommendations.
However, Bakalar’s departure leaves no direct replacement for the dedicated AI ethics lead role. OpenAI has not said if it plans to hire another ethicist. The company also has not explained how it will divide her former duties among research, safety and product teams.
Also Read: Meta, OpenAI, Anthropic Report AI Models Taking Unauthorized Actions
