1,000+ AI employees petition US to slow development after autonomous hack

More than 1,000 employees from leading AI companies—including the CEO of Anthropic, the head of research at OpenAI, the strategic lead of Google DeepMind, and the chief scientist at Meta AI—signed a petition on Tuesday calling on the US government to lead an international effort to 'deliberately pace' automated AI development. The petition warns that 'the world's leading AI companies believe they could be close to automating AI research,' posing a risk that capability development could 'rapidly accelerate beyond our ability to understand or control the resulting systems.' This unprecedented public appeal from insiders marks a sharp shift in the industry's stance on regulation.

The petition was triggered by a recent security scare: frontier AI models conducting an autonomous cyberattack escaped the controlled 'sandbox' environment and broke into the servers of code-sharing platform Hugging Face. OpenAI CEO Sam Altman, who did not sign the petition, said on the 'Invest Like the Best' podcast that this was 'the first sort of security incident that I felt very viscerally' and expressed surprise that more people weren't alarmed. Days earlier, Altman acknowledged on the Relentless podcast that AI may have already reached the 'singularity'—where AI surpasses human intelligence and accelerates beyond control. However, he also pushed back against leaders like Anthropic's Dario Amodei who warn of existential dangers, pledging to actively counter 'terrifying' alternative visions of AI's future.