Sam Altman, the CEO of OpenAI, has resigned from the company’s Safety and Security Committee, a group that was established in May to oversee safety protocols related to OpenAI’s projects and artificial intelligence models.
Going forward, the committee will function independently, chaired by Zico Kolter, a professor at Carnegie Mellon University. Its members include Adam D’Angelo, CEO of Quora; retired U.S. Army General Paul Nakasone; and former Sony Executive Vice President Nicole Seligman—all of whom are also part of OpenAI’s board.
This shift comes in the wake of increased scrutiny from U.S. lawmakers concerning OpenAI’s safety practices and its approach to AI regulation. Altman’s exit follows a situation where five U.S. senators raised questions about OpenAI’s policies, alongside reports indicating that several employees focused on long-term AI safety risks have left the organization.
In addition, OpenAI has ramped up its lobbying efforts in Washington, allocating $800,000 for the first half of 2024, a substantial increase from the $260,000 it spent last year.
The committee recently evaluated OpenAI’s most recent AI model, referred to as o1, after Altman’s departure and will continue to receive ongoing updates on safety evaluations. They maintain the power to postpone the release of AI models until any safety issues are addressed.
In a May opinion piece for The Economist, former board members Helen Toner and Tasha McCauley voiced their concerns regarding the organization’s ability to self-regulate. They stated, “From our experience, we believe that self-governance cannot reliably withstand the pressure of profit motives.”
OpenAI’s financial aspirations are on the rise. The company is reportedly working to secure over $6.5 billion in funding, which could potentially boost its valuation to more than $150 billion. To attract this investment, OpenAI is said to be contemplating a shift away from its existing hybrid nonprofit model, which was originally instituted to limit investor returns and ensure that the mission of developing artificial general intelligence serves humanity’s best interests.