OpenAI has announced that Paul Christiano is joining the board of its nonprofit foundation, as well as its Safety and Security Committee. The appointment brings a well-known figure in AI alignment research into the governance structures that oversee the company's most sensitive deployment decisions.
Christiano is a familiar name within the OpenAI ecosystem. He led the alignment team from 2017 to 2021, contributing to foundational work on reinforcement learning from human feedback (RLHF), a technique that has since become central to training large language models. He later founded the Alignment Research Center (ARC), an independent organization focused on evaluating risks posed by advanced AI systems. More recently, he served as head of safety research at the US government's AI Safety Institute, housed within NIST under the Biden administration.
The appointment comes as OpenAI's governance structure continues to draw scrutiny, with the foundation retaining an oversight role over the company's commercial arm following its restructuring. The Safety and Security Committee, which Christiano now joins, has taken on greater weight since the internal governance tensions of 2023 concerning the pace of model deployment.
Bringing Christiano onto the board can be read as a signal aimed at observers concerned about how quickly generative AI is being developed and shipped. His standing within the AI safety community, where he is regarded as a rigorous and independent voice, could lend credibility to OpenAI's internal risk evaluation processes, though it is unlikely to fully resolve broader questions about balancing nonprofit governance against commercial pressures.