OpenAI announces new security team days after disbanding old one

The company, in a blog post on Monday, announced the formation of the Safety and Security Committee, which it said will be responsible for making recommendations on critical safety and security decisions for all OpenAI projects.

The announcement follows the departure earlier this month of several key members of the security committee, including co-founder Ilya Sutskever and Jan Leike. Leike was particularly critical of OpenAI upon his departure, accusing the company of neglecting “security culture and processes” in favor of “shiny products.” He chose to leave the company, he said, because he “had disagreed with OpenAI’s leadership over the company’s core priorities for some time, until we reached ultimately a breaking point.”

Given these criticisms, Altman’s oversight of the safety committee could come under scrutiny. He will be joined by directors Bret Taylor (who serves as chairman of the board), Adam D’Angelo and Nicole Seligman.

“A first task of the Safety and Security Committee will be to evaluate and further develop OpenAI’s processes and safeguards over the next 90 days,” the company wrote. “At the end of the 90 days, the Safety and Security Committee will share its recommendations with the full Board of Directors. Following the Board’s full review, OpenAI will publicly share an update on the adopted recommendations in a manner consistent with safety and security.

The blog post also reveals that OpenAI is training its “next frontier model,” a successor to the one currently driving ChatGPT, saying that “we anticipate that the resulting systems will take us to the next level of capabilities on our path towards AGI”. .”

News of the new security committee comes less than 10 days after OpenAI disbanded the “Superalignment” team, integrating the remaining members into broader research efforts within the company. Leike and Sutskever were the main members of this superalignment team.