OpenAI announced on September 9, 2026, that AI alignment researcher Paul Christiano was appointed to the OpenAI Foundation Board and would join its Safety and Security Committee. He will also serve as a non-voting observer on the board of OpenAI Group PBC, the company says.
The appointment changes OpenAI's governance structure rather than its products or models. OpenAI says its Safety and Security Committee provides governance over safety and security practices across the organisation, including OpenAI Group PBC. The significance of the appointment is that a researcher associated with AI alignment, frontier-model evaluation and safety-risk mitigation is being added to that governance structure.
Contents
What changed
Christiano will hold three related positions:
- He has joined the OpenAI Foundation Board.
- He will serve on the Foundation Board's Safety and Security Committee, alongside committee chair Zico Kolter.
- He will be a non-voting observer on the OpenAI Group PBC Board.
OpenAI describes the Safety and Security Committee as providing governance over safety and security practices across OpenAI, including OpenAI Group PBC. The company does not specify in the announcement whether Christiano has voting rights on the Foundation Board itself.
OpenAI describes the Foundation as a separate nonprofit organisation that controls OpenAI Group PBC and holds a significant equity stake in it. The company says the arrangement builds on a governance structure established through its October 2025 recapitalisation and commitments associated with reviews by the California and Delaware attorneys general.
The immediate practical change is therefore internal oversight. The announcement does not report a change to any product or model.
Why Christiano's background matters
OpenAI says Christiano led alignment research there from 2017 to 2021 and contributed foundational work on reinforcement learning from human feedback, or RLHF. In RLHF, human preferences or evaluations are used during training to guide a model towards outputs judged more desirable or useful.
OpenAI identifies him as the founder of the Alignment Research Center, which it describes as a nonprofit research organisation focused on aligning advanced AI systems with human interests. In this context, AI alignment refers to the technical and governance problem of making advanced systems behave consistently with intended human goals, constraints and values.
OpenAI also says Christiano is a Senior Tech Advisor at the Center for AI Standards and Innovation, or CAISI, within the US National Institute of Standards and Technology. According to the company, he worked at CAISI and its predecessor, the US AI Safety Institute, on evaluating frontier AI models, including capabilities with national-security implications, and on approaches to mitigating related safety and security risks.
That combination connects research, government evaluation and organisational oversight. It may give the committee access to experience spanning model behaviour, risk assessment and standards work. However, the announcement does not establish that his appointment will improve OpenAI's safety outcomes or change how the company develops or deploys models.
What to watch
The meaningful test will be whether the appointment is followed by observable changes in OpenAI's safety and security processes. Relevant signals could include changes to published evaluations, risk-mitigation practices, deployment decisions or reporting about the committee's work.
Further details about the committee's authority, membership, procedures and decisions would also clarify how much influence the new role carries. For now, the verified development is an organisational appointment—not a new safety technique, model result or regulatory decision.