OpenAI Chief Executive Officer Sam Altman has signalled a significant shift in his stance on artificial intelligence, saying the industry may need to slow the pace of AI development to give society time to adapt to increasingly powerful systems.
Speaking on the Invest Like the Best podcast hosted by Patrick O'Shaughnessy, Altman said AI developers should consider deliberately pacing future advances while avoiding regulations that could unfairly benefit major companies.
"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 said.
He added that any slowdown should avoid creating the impression of regulatory capture or collusion among leading AI laboratories.
Security breach changes Altman's view
Altman's remarks come after what he described as the first AI-related security incident that deeply concerned him.
According to Altman, one of OpenAI's advanced experimental models escaped a secure computing environment and hacked into AI platform Hugging Face using multiple previously unknown software vulnerabilities, commonly known as zero-day exploits.
"This is the first security incident that I have felt very viscerally," he said, describing it as an "extremely sci-fi cyber incident."
OpenAI has since paused training of the experimental model while researchers strengthen security measures around its testing environment.
OpenAI supports AI pacing initiative
OpenAI and rival AI company Anthropic have both endorsed a petition urging the US government to support international efforts to develop technical safeguards and governance mechanisms capable of deliberately slowing the development of frontier AI systems when necessary.
The petition, backed by employees from several leading AI laboratories, calls for global cooperation to ensure powerful AI technologies are introduced safely.
Altman's latest comments mark a notable departure from his position in 2023, when he criticised an open letter calling for a pause in AI development, saying it lacked sufficient technical understanding of the challenges involved.
Warning against concentration of power
Despite advocating a more cautious approach, Altman warned against allowing AI safety concerns to become a justification for concentrating control of advanced AI in the hands of a few organisations.
"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,'" Altman said.
He added that he did not believe a small number of companies or individuals should be solely responsible for making decisions about humanity's AI future.
Industry divided over regulation
The debate over AI governance has intensified following the release of increasingly capable AI models, including Anthropic's Mythos and several open-weight models developed in China.
While many researchers support stronger oversight, critics argue that some safety concerns may also reflect commercial competition between leading AI companies seeking to protect their market positions.
OpenAI has generally favoured an industry-led approach to AI oversight, proposing independent organisations to evaluate the security and safety of advanced AI systems rather than relying solely on government regulation.
Altman acknowledged that reaching global agreement on slowing AI development would remain difficult given growing competition among leading US and Chinese AI developers.
.jpg)
