Beijing warns of six AI risks, rejects US slowdown calls

Beijing's warning came from an unusual corner: the minister of state security, writing in a weekend essay that framed AI as a direct challenge to the political regime, digital infrastructure, public order, and national defense. That framing is notable because it elevates AI from a commercial or scientific topic to a core security concern.

At the same time, Beijing is not slowing down. It rejected US calls to do so, and the state-backed Global Times attacked Anthropic's CEO for suggesting a slowdown, calling his argument part of the "Cold War playbook." The body pairs these two positions: anxiety about AI's risks, and refusal to limit its development.

Some Chinese analysts, the article notes, see a way out through diplomacy. They argue Beijing should push for a superpower compact on AI and to "advance through concessions," a hedge aimed at retaining some sway over global regulation and access to key international markets. Whether that view gains traction in Beijing is likely to shape how AI rules are set internationally, and it is the piece of this story that remains undecided.