Anthropic has halted global access to its Fable 5 and Mythos 5 models following a direct government order issued on Friday evening. The directive required the company to block usage for all foreign nations, including its own employees, citing national security concerns. In response, Anthropic immediately severed connections for all customers worldwide. The firm stated it was complying with the mandate but noted that the government provided no specific details regarding the nature of the security threat. Anthropic claimed that any evidence of potential jailbreaks was shared verbally and described the identified vulnerabilities as minor, suggesting they were already present in other available models.
This incident highlights the increasing friction between artificial intelligence developers and state security apparatuses. By cutting off access to advanced models without public justification, the government sets a precedent for arbitrary restrictions that could stifle innovation. The lack of transparency regarding the specific risks undermines trust between regulators and technology companies. Furthermore, the decision to include internal staff in the blackout suggests that the perceived threat extends beyond external users to the very personnel who build and maintain these systems. This move signals a tightening of control over frontier AI capabilities, potentially forcing companies to self-censor or discontinue features deemed risky by undefined state criteria.
- Anthropic blocked Fable 5 and Mythos 5 globally after a government order citing unspecified national security concerns.
- The company received no written details from officials regarding the specific vulnerabilities or threats identified.
- Internal employees and external customers alike lost access to the models, indicating a broad suspension of usage.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
