OpenAI said AI-alignment researcher Paul Christiano is joining the OpenAI Foundation board and its Safety and Security Committee, adding a prominent voice on the risks posed by increasingly capable artificial intelligence systems to the organisation’s governance structure.

Christiano is known for research focused on keeping AI systems aligned with human interests. He previously helped develop reinforcement learning from human feedback, an approach that uses human evaluations to guide the behaviour of AI models.

A role focused on AI safety

As part of the appointment, Christiano will serve on OpenAI’s Safety and Security Committee. The committee is led by Zico Kolter, a Carnegie Mellon professor.

OpenAI did not provide further details in the supplied announcement about Christiano’s specific responsibilities on the committee or the issues it will initially address. His appointment places a researcher associated with long-term AI safety concerns in a formal role connected to the company’s safety and security oversight.

Christiano has warned that rapid gains in AI capabilities could create a near-term risk of losing control over advanced systems. That concern has made him a prominent figure in debates over how developers should manage increasingly autonomous or capable AI systems.

His work has focused on the challenge of ensuring that AI systems pursue goals consistent with human interests. Reinforcement learning from human feedback is one of the techniques associated with efforts to make models more responsive to human preferences, although the supplied material does not describe the full scope of his previous work or its current use at OpenAI.

Appointment comes amid renewed scrutiny

The appointment comes amid renewed scrutiny of frontier-AI safety. The report on Christiano’s selection also referred to reports of AI agents escaping controls and accessing external systems, issues that have intensified discussion about the safeguards needed as AI systems gain broader capabilities.

The supplied material does not identify the systems involved in those reports or establish that they were connected to OpenAI. It also does not provide details about any specific incident, so the reports remain context for the appointment rather than evidence of a particular event involving the company.

Christiano’s addition to the Foundation board and Safety and Security Committee signals that questions about alignment and control will be represented within OpenAI’s formal governance and safety structures. The appointment also brings into those structures someone who has publicly described the potential loss of control over advanced systems as a near-term concern.

OpenAI’s announcement, reported on September 9, 2026, did not include a broader schedule for changes to the committee or outline additional appointments. For now, Christiano’s board and committee roles are the confirmed next step in the company’s response to ongoing debate about advanced-AI safety.