“If this standard was applied across the industry, we believe it would essentially halt all new model deployments for all frontier model providers.”

AnthropicAn AI safety company — creators of the Claude family of language models. complied with the export-control order but disputed the rationale: the reported bypass was narrow, and other frontier models showed similar capability. The pushback previewed an industry fight over when a jailbreak should trigger withdrawal versus patched safeguards.

AnthropicAn AI safety company — creators of the Claude family of language models.'s counter-argument to the first frontier-model recall order.

It defined the stakes for every lab shipping cyber-capable models after Fable.