OpenAI has experienced a series of high-level exits from its safety and ethics ranks in a short span of time. Chloé Bakalar, who served as the company’s head of ethics, stepped down less than twelve months after she arrived. Her departure follows those of Johannes Heidecke, who headed safety systems, and Joshua Achiam, previously responsible for mission alignment and later holding the title of chief futurist.
Bakalar joined OpenAI around August 2025 after a lengthy period at Meta, where she worked as chief ethicist and helped embed ethical considerations into major products.
At OpenAI, her focus included the ethical dimensions of building advanced models, interactions between humans and AI systems, and related philosophical questions such as machine consciousness.
Reports indicate she left in July 2026 without a public announcement from the company, and no immediate successor for the dedicated ethics role has been named.
These moves form part of a wider pattern of turnover in OpenAI’s safety-related leadership.
Heidecke’s exit occurred amid an internal reorganization that brought safety teams under research leadership, with expanded responsibilities assigned to the vice president overseeing research and alignment.
Achiam, a long-tenured researcher who joined as an intern years earlier and later led mission alignment efforts before that unit was dissolved, also announced his departure around the same period after nearly a decade with the organization.
The timing of these leadership changes coincides with intensified external attention on how leading AI developers manage risks associated with increasingly capable systems.
Concerns continue to grow around model control, the effectiveness of human oversight mechanisms, and the balance between rapid capability advances and robust safeguards.
OpenAI has faced questions about incidents during testing of advanced systems and has taken steps such as adjusting timelines for certain model work while reinforcing internal security practices.
Broader discussions among policymakers and industry participants have focused on appropriate controls for AI technologies, including potential restrictions on the release of highly capable models.
Industry professional now generally note that OpenAI has undergone multiple structural shifts involving its safety functions in recent years, including the creation and later dissolution of specialized teams focused on alignment and superalignment.
Earlier departures of other safety-oriented researchers and leaders have similarly drawn attention, with some individuals moving to rival organizations or independent initiatives centered on AI safety standards.
The company has emphasized efforts to more tightly integrate safety considerations into core research and product decisions rather than maintaining fully separate reporting lines.
The absence of a dedicated ethics lead at this stage leaves OpenAI without a single senior figure solely focused on that domain at a moment when ethical and safety questions feature prominently in public and regulatory conversations about frontier AI development.
Industry professionals now continue to monitor how the organization addresses governance, risk mitigation, and long-term alignment challenges as it advances its most powerful models.
As first reported by the FT, these developments underscore ongoing tensions within the AI sector between accelerating technological progress and ensuring adequate oversight and ethical frameworks. As scrutiny from governments, researchers, and the public remains elevated, leadership stability in safety and ethics roles at major labs is likely to stay under close observation.
