Anthropic is urging governments and leading AI companies to agree to a global freeze on developing more powerful AI systems, warning that unchecked progress could lead to loss of control. According to multiple reports, the company says it would halt work on further advanced models if competitors in the United States and China make a similar commitment at the same time. The proposal is presented as an attempt to reduce the risk that one side accelerates development while others pause, creating a strategic imbalance.
All three outlets report that Anthropic’s offer is conditional on coordinated action by rivals, reflecting the company’s view that unilateral restraint is unlikely to be effective. The warning focuses on the possibility that increasingly capable systems could introduce safety and governance challenges that are difficult to manage once they are widely deployed.
The reports do not indicate that any government or company has agreed to the freeze, but they describe the proposal as an effort to prompt international discussion and cooperation among major AI players.
