Government Orders Anthropic to Pull Claude Fable, Mythos AI Models

The U.S. government issued an emergency export control directive on Friday ordering Anthropic to immediately suspend access to its two most powerful AI models Claude Fable 5 and Claude Mythos 5 for all foreign nationals, including the company’s own employees. The directive applies inside and outside the United States and forced Anthropic to disable the models for its entire customer base to ensure compliance. Officials said the vulnerabilities prompting the action were relatively straightforward to exploit, while Anthropic contended the finding was not severe and noted that other publicly available models could reveal similar weaknesses without bypassing guardrails. The government has not disclosed full technical details of the concern, but the action underscores the administration's willingness to curb frontier AI deployment on national security grounds. The move also prompted public and political discussion about safety versus rapid deployment, as the company noted access to other models remains unaffected and pledged to restore access as quickly as possible. On Saturday, David Sacks claimed a jailbreak was demonstrated to a tester, a claim Anthropic disputed as overstated, while reiterating that the incident should not halt broader AI progress.