THE U.S. government has implemented an emergency export control directive banning the Claude Fable 5 AI model from being offered to non-U.S. citizens due to critical security vulnerabilities. Anthropic, the model's developer, is in negotiations with the government to resolve the issue, which revolves around jailbreak vulnerabilities that could allow users to bypass safety protocols.
Despite Anthropic's insistence that the government exaggerates the risks, the government remains firm in its position, demanding a comprehensive resolution to the jailbreak issues before lifting the export ban. Negotiations have stalled, leaving the future availability of Claude Fable 5 uncertain.