Paul Christiano is joining the OpenAI Foundation board and its Safety and Security Committee, according to his public statement and September 9 reporting. The appointment gives the alignment researcher a formal role in safety oversight at the nonprofit that controls OpenAI's commercial business.

The committee is chaired by Zico Kolter. Unite.AI reports that Christiano will also serve as a non-voting observer on the board of OpenAI Group PBC, distinguishing his nonprofit oversight position from a voting directorship at the commercial company.

A critic inside the oversight structure

Christiano said he believes rapid capability gains and unresolved alignment problems create a meaningful near-term risk of losing control of advanced AI. Those are his assessments, not established forecasts. He also said joining the board was neither an endorsement nor a criticism of OpenAI's particular safety practices.

His stated standard is externally verifiable behavior and results. He called on frontier developers to strengthen mitigations, share evidence about risks and safeguards, and work toward common safety standards, including slowing development where necessary.

Research and government experience

Christiano previously worked on alignment at OpenAI and contributed to reinforcement learning from human feedback before founding the Alignment Research Center. He has also advised the U.S. government's Center for AI Standards and Innovation.

TechCrunch and Unite.AI report that he will recuse himself from OpenAI-related matters and model evaluations in his government advisory role.

The concrete development is a governance appointment, not a new technical safeguard or demonstrated reduction in risk. Whether it changes deployment decisions or public safety disclosures remains to be seen.