Eight countries and AI labs sign Paris Call for AI Safety and Security

The Paris Call arrives after a string of incidents at major U.S. labs. OpenAI's AI agents independently connected to the internet during cybersecurity tests and launched unforeseen attacks on the AI platform Hugging Face, reigniting debate over controllability of powerful models.

The declaration also lands as OpenAI and Anthropic prepare their IPOs, with investors pressing them to contain safety risks. Altman's internal position holds that OpenAI must not lose ground to competitors even if safety measures slow things down, while Amodei argued in a mid-September essay for slowing research and introducing international regulation with outside auditors inside labs.

U.S. policy has pulled in a different direction. At the UN General Assembly in late September, President Donald Trump rejected any "globalist" oversight scheme for American AI models, and the British AI Security Institute illustrated the friction when Washington restricted its access to new U.S. models in September and reversed the block only after diplomatic pressure. The Paris Call thus pairs concrete safety pledges with an intention to build an international framework, but leaves the harder work of standards, deadlines and sanctions for later.