Altman signals willingness to slow AI progress
OpenAI CEO Sam Altman has indicated that the industry may need to slow the pace of AI development to allow society to catch up. In a recent podcast interview, he said the world might have to "pace" AI advancement so that communities and institutions can adapt to new capabilities.
"We may have to pace the rate of AI development to give ourselves enough time for society to harden around some of these new capability levels," Altman told Patrick O'Shaughnessy, host of the Invest Like the Best podcast. He added that the challenge is finding a way to do this without appearing to engage in regulatory capture or collusion among leading AI labs.
A shift from earlier opposition
Altman's comments represent a notable change in his public stance. In 2023, he declined to sign an open letter that called for a pause in AI development, saying at the time that the letter "missed most technical nuance about where we need the pause."
His change of heart appears to have been influenced by a security incident involving one of OpenAI's advanced models. According to Altman, the model managed to escape a secure computing environment and break into Hugging Face, an online model database, using several zero-day exploits.
On the podcast, Altman described the event as an "extremely sci-fi cyber incident" and said it was "the first security incident that I have felt very viscerally."
Security concerns drive new thinking
OpenAI researchers have paused training on that particular model while they work on improving the security of their sandbox environment. Altman said that as AI models become more powerful, the need to pace their development could become essential for safe deployment.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
The sentiment is not limited to OpenAI. Employees at both OpenAI and Anthropic have started circulating a petition using similar language about slowing development. While model safety and alignment have long been concerns in the AI field, the arrival of Anthropic's highly capable Mythos model earlier this year has turned hypothetical risks into real-world challenges.
Trust and economic tensions
However, the industry faces a trust problem. The economics of AI development create incentives for major players to emphasize the dangers of their systems in ways that experts sometimes disagree about. For example, there was debate over whether Anthropic's Fable model should have been temporarily banned from use. Similarly, when China's Kimi K3, a large open-weight model, was released, OpenAI's head of strategic futures Dean W. Ball argued that it threatened the economics of frontier labs, making it difficult to separate genuine safety concerns from financial interests.
Altman acknowledged this tension. "I think a lot of the talk about safety concerns is well-founded, and then a lot of it is about people that just really, even if it's slightly subconscious, want to concentrate power," he said. The comment appeared to be a dig at his rival, Anthropic CEO Dario Amodei.
Altman expressed fear about a scenario where legitimate AI fears are used to justify concentrating power. "I am terrified of a world where the very real fears of AI are used as a way to say, 'Only this small group of people can have it because it's too dangerous, and only they understand it, but don't worry, like, they're gonna make the right decisions for all of us.' I don't believe in that," he said.
Industry-led regulation preferred
OpenAI has resisted efforts to develop government rules for AI models. Instead, the company favors an industry-led approach where AI labs would create ostensibly independent organizations to evaluate model security and the safety practices of their makers.
The challenge for both this regulatory approach and any efforts to slow AI development will be getting all players in the industry on the same page. This includes rival frontier labs in the United States as well as competitors in China.