Sam Altman Steps Down from OpenAI’s Safety Board, Citing Shift to Independent Oversight

Sam Altman, the CEO of OpenAI, is stepping down from the internal commission he founded in May to supervise “critical” safety choices about the business’s operations and projects.

The Safety and Security Committee, which is chaired by Carnegie Mellon professor Zico Kolter and includes Quora CEO Adam D’Angelo, retired U.S. Army General Paul Nakasone, and former Sony EVP Nicole Seligman, will now function as an “independent” board monitoring body, according to a blog post published by OpenAI. They are all currently serving on OpenAI’s board of directors.

After Altman resigned, the commission reviewed the safety of OpenAI’s most recent AI model, o1, according to a post by the company. According to the business, the group will still get frequent updates from OpenAI’s safety and security teams. It will have the authority to postpone releases until any safety issues are resolved.

Also Read: Microsoft AI Chief Mustafa Suleyman Praises OpenAI’s Sam Altman, Discusses Future of AI and Regulation

OpenAI stated in the post that “as part of its work, the Safety and Security Committee… will continue to receive regular reports on technical assessments for current and future models, as well as reports of ongoing post-release monitoring. [We] are expanding on our current model launch procedures and practices to create a comprehensive safety and security framework with precisely defined success standards for model launches.”

After five US senators questioned OpenAI’s principles in a letter to Altman this summer, Altman resigned from the Safety and Security Committee. Ex-OpenAI researchers have accused Altman of resisting “real” AI legislation in favor of measures that further OpenAI’s corporate goals. Nearly half of the OpenAI personnel that previously concentrated on the long-term risks associated with AI have departed.

To their point, OpenAI has allocated $800,000 for the first half of 2024 compared to $260,000 for the entire previous year, a significant increase in the company’s spending on federal lobbying. In addition, Altman became a member of the Artificial Intelligence Safety and Security Board earlier this spring, offering suggestions for advancing and applying AI in vital infrastructure across the United States.

Also Read: Top 50+ Sam Altman Quotes on AI for Great Products and Future Technology

There is little evidence that the Safety and Security Committee would make tough choices that materially affect OpenAI’s commercial roadmap, even in Altman’s absence. Remarkably, OpenAI declared in May that it would try to address “legitimate criticisms” of its work through the commission; of course, “legitimate criticisms” are subjective.

Former OpenAI board members Helen Toner and Tasha McCauley stated in a May opinion piece for The Economist that they did not believe OpenAI as it is now capable of holding itself accountable. “[We] feel that self-governance cannot consistently withstand the pressure of profit incentives based on our experience,” they stated in their letter.

Moreover, OpenAI’s financial incentives are rising.

OpenAI is believed to be valued at over $150 billion, and the business is reportedly in the process of raising over $6.5 billion in a capital round. According to reports, OpenAI may give up on its hybrid nonprofit corporate structure to seal the agreement. This structure attempted to limit investor returns in part to make sure OpenAI stuck to its original goal of creating artificial general intelligence that “benefits all of humanity.”

Also Read: Sam Altman Suggests “Universal Basic Compute” to Enable Worldwide Access to GPT-7’s Assets