OpenAI Names RLHF Pioneer Paul Christiano to Its Safety Oversight Board

News Summary

OpenAI has appointed Paul Christiano, a researcher long associated with warning about catastrophic risks from advanced artificial intelligence, to its nonprofit Foundation board and to the Safety and Security Committee that oversees model release decisions across the company. The announcement was made on September 9, 2026 (Pacific Time), and it places one of the field's most prominent risk-focused voices inside the governance structure of the company building some of the world's most capable AI systems.

Who Paul Christiano Is

Christiano is best known as a co-developer of reinforcement learning from human feedback (RLHF), the training technique that underlies how modern large language models are aligned to follow instructions and produce helpful, well-behaved responses. He worked at OpenAI from 2017 to 2021, leading alignment research during that period, before leaving to found the Alignment Research Center (ARC), a nonprofit dedicated to developing theoretical and empirical tools for keeping advanced AI systems aligned with human intentions.

Since leaving OpenAI, Christiano has also taken on a government advisory role. He serves as a Senior Technical Advisor at the Center for AI Standards and Innovation (CAISI), housed within the U.S. National Institute of Standards and Technology (NIST), where he has contributed to the evaluation of frontier AI models for safety-relevant capabilities. According to OpenAI, this experience spans work across two administrations, giving him a vantage point on AI safety that blends technical research with policy and standards development.

The New Role at OpenAI

Under the appointment, Christiano joins the OpenAI Foundation board and, more specifically, its Safety and Security Committee (SSC), working alongside the committee's chair, Carnegie Mellon University professor Zico Kolter. The SSC is the body responsible for governance over safety and security practices across OpenAI, including the OpenAI Group public benefit corporation, and it holds the authority to review and sign off on the readiness of new frontier models before they are released to the public.

Christiano's position on the OpenAI Group PBC board is that of a non-voting observer, and OpenAI has said that in his advisory capacity he recuses himself from matters and model evaluations tied to his separate government work at CAISI, a structure intended to avoid conflicts of interest between his public-sector and private-sector roles.

Why the Appointment Stands Out

The appointment is notable because Christiano has been outspoken about the scale of risk he associates with the pace of AI development. In a statement accompanying the announcement, he wrote that he now believes there is "a meaningful risk that rapid acceleration in AI capabilities leads to catastrophic and irreversible loss of control in the very near term," and that he does not believe the AI industry as a whole, including OpenAI, is currently on track to reduce that risk to an acceptable level. He added that he chose to join because he believes that if OpenAI "rises to the occasion," the company could meaningfully reduce that risk.

Bret Taylor, chair of the OpenAI Foundation board, framed the appointment as a strengthening of independent oversight, saying that Christiano "has helped define the field of AI alignment through work that is rigorous and focused on the hardest questions posed by increasingly capable systems," and that his experience in frontier AI safety and government standards work would strengthen both the board's oversight and the committee's work. Christiano himself noted that "AI capabilities have advanced very rapidly in the last year and alignment remains a difficult technical problem," underscoring why he sees the committee's responsibility as more important now than at any prior point.

Context Within OpenAI's Governance Structure

The move comes after OpenAI's 2025 restructuring, which separated its original nonprofit Foundation from the newly formed OpenAI Group public benefit corporation, a structure designed to keep a nonprofit mission and safety oversight layer above the commercial entity. The Safety and Security Committee already included figures such as Zico Kolter, Adam D'Angelo, Sue Desmond-Hellmann, Paul M. Nakasone, and Nicole Seligman, alongside CEO Sam Altman. Adding a researcher known for emphasizing existential-scale risk broadens the range of perspectives represented on the committee that has final input on whether new models are safe enough to ship.

Broader Industry Backdrop

Christiano's appointment lands at a moment when questions about AI safety governance are drawing wide attention across the technology sector. Public debate over how quickly capable AI systems are advancing, and how well safety research is keeping pace, has intensified industry-wide in 2026, with researchers across multiple labs publishing work on alignment, interpretability, and containment. By bringing a researcher with both deep technical alignment credentials and government standards experience into a formal oversight role, OpenAI is signaling that it wants its safety governance to include voices who have been willing to publicly state that current industry practices are not yet sufficient, rather than only those aligned with a more optimistic timeline.

For students and observers following the AI safety field, the appointment offers a concrete example of how alignment research — once a niche academic pursuit — has moved into corporate governance, with researchers who study how to keep AI systems predictable and controllable now sitting on the boards that decide when a new model is ready for the world.