OpenAI Adds Safety Critic Paul Christiano to Foundation Board

OpenAI named Paul Christiano to the board of the OpenAI Foundation on Sept. 9, 2026, five years after he left the company's alignment research team to found a nonprofit warning that AI labs are not on track to control the systems they are building.

TechCrunch reported the appointment the same day, citing OpenAI's own announcement. Christiano joins the Foundation's Safety and Security Committee, chaired by Zico Kolter, a Carnegie Mellon University professor, and will also sit as a non-voting observer on the board of OpenAI Group PBC, the company's for-profit subsidiary. The committee holds final authority over whether OpenAI releases a new model, according to TechCrunch's report.

"I now believe there is a meaningful risk that rapid acceleration in AI capabilities leads to catastrophic and irreversible loss of control in the very near term," Christiano wrote in the post announcing his appointment, TechCrunch reported. "I do not think that the AI industry in general, including OpenAI, is currently on track to reduce this risk to an acceptable level."

Christiano's path back to OpenAI, since 2017

Christiano led alignment research at OpenAI from 2017 to 2021 and contributed to the development of reinforcement learning from human feedback, a technique used to steer large language models like OpenAI's GPT-6 Astra toward following instructions rather than just predicting text. He left in 2021 to found the Alignment Research Center, a nonprofit that studies whether advanced AI systems could act against the interests of the people who built them.

Period	Role	Organization
2017-2021	Led alignment research	OpenAI
2021-present	Founder	Alignment Research Center
Ongoing	Senior technical adviser	Center for AI Standards and Innovation (NIST)
Sept. 9, 2026	Foundation Board, Safety and Security Committee	OpenAI

Christiano also serves as senior technical adviser at the U.S. Center for AI Standards and Innovation, a National Institute of Standards and Technology unit that evaluates frontier models for national security risk, according to Unite.AI's coverage of the appointment. He will recuse himself from OpenAI-related reviews in that government role, Unite.AI reported.

What the Safety and Security Committee controls

The Foundation board Christiano is joining controls OpenAI Group PBC through a 26 percent equity stake and special voting rights, Unite.AI reported, a structure put in place when the company completed its recapitalization in October 2025. Kolter has chaired the Safety and Security Committee since that restructuring, according to Unite.AI. Christiano is the panel's most prominent addition since then.

Neither TechCrunch's report nor Unite.AI's coverage specifies whether Christiano will see model evaluation results before a release goes out, or whether the committee's decisions are made public in any form. OpenAI's own Agents API beta launched this month without any public sign of Safety and Security Committee review, and it is not clear from either report whether that will change now that the committee has a member on record saying the industry is not managing its risk to an acceptable level.

Unite.AI put the Foundation's 26 percent stake in OpenAI Group PBC at a valuation of roughly $130 billion, which makes the committee Christiano just joined a check on the pace at which the company's single most valuable asset ships new models.