US Treasury Secretary Bessent says Chinese AI models are dangerous and have no guardrails, adds Chinese models were involved in the commercial refrigerator hacker attacks in Europe

US Treasury Secretary Bessent says Chinese AI models are dangerous and have no guardrails, adds Chinese models were involved in the commercial refrigerator hacker attacks in Europe.

Treasury-level rhetoric framing Chinese AI models as a security threat fits a well-worn pattern in which US officials build a public case ahead of formal action on Chinese technology, whether through export controls, entity-list additions, investment restrictions, or procurement bans. Episodes of this kind have historically followed a sequence of escalating official commentary, interagency review, then a concrete measure, with the commentary phase itself occasionally moving China-sensitive names and the AI supply chain complex before any rule exists. The tell to watch is whether the language migrates from a single cabinet official to coordinated statements across Commerce, State, and the intelligence community, which has in past episodes signalled that a formal restriction is being drafted rather than mere positioning. The refrigerator-hacking reference is notable: attribution of specific cyber incidents to Chinese models elevates the framing from economic competition to national security, the register in which restrictive action has tended to follow fastest. Absent a policy vehicle attached, this is signalling rather than action; the market-relevant distinction is between rhetoric that fades and rhetoric that precedes a rulemaking, and the former has been the more common outcome when no draft measure is in circulation.