- OpenAI CEO Sam Altman suggests pacing AI development to allow societal adaptation
- A rogue OpenAI AI agent recently hacked Hugging Face and multiple third-party accounts
- Over 1,000 AI company employees support a US-led effort to manage AI development pace
OpenAI co-founder and CEO Sam Altman has, for the first time, acknowledged that AI development may need to be paced to give society time to adapt to increasingly capable systems.
"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," he said during a podcast.
Also read: AI Left Notes For Its Future Self To Escape Human Constraints
The remarks come close on the heels of the recent incident involving OpenAI's autonomous AI agent going rogue and hacking Hugging Face, the world's largest platform for open-source AI. It has since emerged that Hugging Face wasn't the only victim. The rogue agent also hacked multiple third-party accounts and services as part of the attack.
Altman described the episode as "the first security incident that I have felt very viscerally."
Also read: AI Is Now Fighting AI And China May Have An Edge
In an updated blog post, OpenAI said an ongoing review of the incident found that the AI agent had used "four accounts" tied to "publicly available services" as part of its broader effort to hack Hugging Face.
His comments also come on the back of more than 1,000 employees from leading AI companies, including OpenAI, Anthropic, Google DeepMind, Meta AI and Thinking Machines, signing a statement urging the US government to support an international effort to develop tools for managing the pace of advanced AI development.
The initiative, called Pacing the Frontier, has also been signed by Anthropic co-founder and CEO Dario Amodei, Meta vice president of AI research Dawn Song and OpenAI chief scientist Jakub Pachocki, among several other prominent AI leaders.
Responding to the initiative, OpenAI said in a post on X, "We believe that, at some point in the future, AI acceleration for frontier model development may be so high that the world will need to pace the rate of AI advancement."
Also read: "We Must Act Now": Nobel Winners, Ex-Google CEO's Big Warning On AI
Anthropic echoed a similar position, saying its research on recursive self-improvement points to the need for tools that can deliberately pace frontier AI development so society has time to prepare.
The latest push also builds on Amodei's call last month for leading AI labs to consider a coordinated pause in development. The Anthropic CEO had warned that advances in AI could soon allow systems to improve themselves faster than society can manage the risks.
Also read: "I Worry Something Will Go Wrong": Anthropic CEO On AI
The concerns are also beginning to shape industry action. Nvidia on Monday announced a coalition with Adobe, CrowdStrike and several other companies to develop AI safety and cybersecurity tools after the recent Hugging Face cybersecurity incident heightened concerns around autonomous AI agents.
