Anthropic, the AI company behind the Claude family of models, is sending senior staff to meet with the Trump administration today in an attempt to settle an escalating dispute over its Mythos AI systems.
At the center of the conflict is a US directive issued on June 13, 2026, that ordered Anthropic to suspend access to its Claude Mythos 5 and the related Fable 5 model for foreign nationals. The reason: a jailbreak vulnerability that the government says could pose national security risks.
What happened and why it matters
Anthropic launched Mythos 5 and Fable 5 in June 2026, positioning them as a leap forward in AI capability. Almost immediately, the government stepped in with an emergency export control directive, citing concerns that the jailbreak vulnerability could be exploited by foreign actors.
Anthropic complied with the order, restricting access as directed. Anthropic has publicly contested the directive, arguing that jailbreak vulnerabilities are not unique to its systems.
