Alignment Research Center founder joins OpenAI board

Paul Christiano, founder of the Alignment Research Center, is joining the OpenAI Foundation Board and taking a seat on its Safety and Security Committee, the body that provides governance over safety and security practices across OpenAI. The appointment was announced by OpenAI, which also placed him as a non-voting observer on the OpenAI Group PBC Board.

At a glance

• The Safety and Security Committee provides governance over safety and security practices across OpenAI, and Christiano's seat on it comes with a place on the Foundation Board that oversees the wider structure.
• OpenAI points to Christiano's work on AI alignment and his years at NIST, the US standards agency, as the background that will strengthen the Foundation's oversight and its scrutiny of internal safeguards.
• On the OpenAI Group PBC Board his role is observational rather than voting, giving Christiano visibility into the for-profit arm's deliberations without a formal say in its decisions.

The appointment reads as an attempt to place an external safety credential inside OpenAI's governance at a moment when the nonprofit Foundation's authority over the for-profit arm is the most scrutinised part of the structure. Christiano's profile, alignment research combined with government standards work, likely matters more for the signal it sends to regulators and researchers than for any single decision the committee will take.

OpenAI framed the move in terms of oversight capacity: as AI capabilities advance, safety, security, alignment and governance matter more than ever, and Christiano is expected to bring "an independent voice to challenge assumptions, assess safeguards, and reinforce accountability around critical decisions."

Before the appointment Christiano founded the Alignment Research Center and worked for years at NIST, the US National Institute of Standards and Technology. Alignment research, the field he is identified with, concerns making model behaviour match the intentions of the people building and deploying the systems. The Financial Times headlined the move as OpenAI naming a top US official from the model testing agency to its board.

The OpenAI Foundation is the nonprofit that sits atop the structure and governs OpenAI Group PBC, the public benefit corporation running the commercial business. The Safety and Security Committee operates on the nonprofit side of that split, with a remit covering safety and security practices across the whole organisation.

Start date and committee roster

The announcement does not give a start date, name the other members of the Safety and Security Committee, or state how long the board term runs. It also does not spell out whether the non-voting observer role on the OpenAI Group PBC Board carries any right to review specific model releases. How much the seat changes committee practice will only become visible in future safety decisions.