PITTSBURGH (TNS) — As OpenAI grapples with its next major leadership shakeup, the $80 billion ChatGPT maker has added Zico Kolter, a top researcher at Carnegie Mellon University, to its board.
Kolter, the newly named director of CMU’s machine learning department, will bring technical and AI safety expertise to OpenAI’s board of directors and will serve on the safety and security committee, the company announced Thursday.
“I’m looking forward to sharing my perspectives and expertise on AI safety and robustness to help guide the amazing work being done at OpenAI,” Kolter wrote on X.
“As we build bigger and bigger systems, we’re going to need more and more automated ways to assess their safety,” he told the Post-Gazette shortly before boarding a plane to meet with his fellow board members.
His appointment comes the same week three top leaders left OpenAI, including co-founders John Schulman and Greg Brockman.
Schulman, who was part of the same safety committee that Mr. Kolter joined, is leaving to join rival Anthropic. Brockman, the company’s president, is taking a sabbatical. Another co-founder, Ilya Sutskever, left the company earlier this year. The company’s board has seen similar volatility ever since CEO Sam Altman was forced out and rehired last fall.
Kolter’s appointment should balance criticism that OpenAI’s safety committee, created in May, was made up primarily of internal staff. He is also the only the AI researcher on OpenAI’s board.
In a statement, OpenAI board chairman Bret Taylor said “Zico adds deep technical understanding and perspective in AI safety and robustness that will help us ensure general artificial intelligence benefits all of humanity.”
OpenAI research scientist and CMU alum Noam Brown praised his former mentor on X, writing that Kolter is “someone I’d frequently turn to for research and career advice.”
“He’s loved by his students and is a world expert in machine learning. I’m thrilled that he’s joining us!”
In an interview, Brown said Kolter’s deep technical expertise should help him understand the vulnerabilities that AI systems face. As a researcher and educator, Kolter is “very quick to adapt to new paradigms,” Brown said.
Before joining CMU in 2012, Kolter earned his doctorate in computer science from Stanford University and then completed a postdoctoral fellowship at the Massachusetts Institute of Technology. He currently serves as chief expert at Bosch and chief technical advisor of the Pittsburgh startup Gray Swan AI. He was appointed to lead CMU’s machine learning department in June.
Last fall, he and the team behind Gray Swan proved it was possible to hack OpenAI’s flagship product, ChatGPT, getting the chatbot to spit out insults and bomb recipes.
“You can’t rely on companies to do all their own auditing,” Kolter told the Post-Gazette at the time.
