Anthropic suspends access to its Claude “Fable 5” and “Mythos 5” AI models amid security concerns connected to U.S. government action, according to multiple reports. Euronews says Anthropic states it believes the U.S. government has identified a potential way to “jailbreak” Fable 5, prompting the company to halt access to the affected models. Other outlets report that the stop follows U.S. restrictions that bar Anthropic from providing the models to foreign users, framing the decision as a compliance and safety measure. Coverage also notes that the development is drawing political and industry attention in Europe, with reactions described as a “wake-up call” for how AI deployment and access controls may change across jurisdictions. The reports collectively indicate that Anthropic still retains control over the models’ availability in line with the restrictions, but access for international users is paused while security issues and regulatory conditions are addressed. The situation highlights the interaction between AI platform governance, cybersecurity risk assessment, and national security-related export or access limits.