OpenAI puts longtime alignment researcher Paul Christiano on its safety board

OpenAI has appointed AI alignment researcher Paul Christiano to the OpenAI Foundation Board and its Safety and Security Committee.

Christiano will also become a non-voting observer on the board of OpenAI Group PBC.

The appointment puts a longtime researcher focused on advanced AI risk directly inside the governance structure responsible for overseeing OpenAI's safety and security practices.

Who is Paul Christiano?

Christiano previously led alignment research at OpenAI from 2017 to 2021.

His work contributed to early research on reinforcement learning from human feedback, commonly known as RLHF, a technique that became important in the development of modern conversational AI systems.

He later founded the Alignment Research Center, a nonprofit focused on ensuring advanced AI systems remain aligned with human interests.

Christiano has also worked with the US government's AI safety efforts and currently serves as a Senior Technical Advisor at NIST's Center for AI Standards and Innovation.

OpenAI says his government role has included evaluating frontier models and studying capabilities with national security implications.

OpenAI is bringing a critic of current safeguards closer to governance

What makes the appointment particularly interesting is Christiano's position on AI risk.

OpenAI's own announcement describes him as someone who takes seriously the possibility that advanced AI systems could create catastrophic risks.

The company also describes him as an independent voice on whether the industry's existing safeguards are adequate.

Christiano said rapid capability advances have made alignment increasingly difficult and strengthened the importance of the Safety and Security Committee.

The committee is responsible for governance of safety and security practices across OpenAI, including OpenAI Group.

The appointment does not resolve the broader debate over whether AI companies can effectively oversee their own frontier systems.

But it places one of the researchers most associated with that debate closer to OpenAI's highest-level safety decisions.