US blocks access to Anthropic's Claude AI after jailbreak flaws

The U.S. government has issued an export control directive affecting Anthropic's Claude Fable and Mythos 5 AI models due to national security concerns, leading to a suspension of foreign access to these technologies. This move comes in light of vulnerabilities identified in the Claude Fable architecture, which have allegedly been exploited through jailbreaking techniques.

Anthropic has expressed frustration over the lack of evidence supporting the government's decision and argues that it impacts a significant user base unfairly. The company is engaged in discussions with officials to potentially restore access.