According to reports highlighted by Android Authority, the U.S. government has ordered Anthropic to suspend access to its most advanced AI models, Fable 5 and Mythos 5, for foreign nationals due to national security concerns. Anthropic said it received an export-control directive and immediately began disabling access, even though the company claims it was not given detailed evidence supporting the government's concerns.
The government's reported concern centers on the possibility that the models could be "jailbroken" to help identify software vulnerabilities, potentially making them useful for cyberattacks. Anthropic disputes this assessment, arguing that the alleged vulnerabilities are minor, already known, and do not provide capabilities beyond those available in other leading AI systems. The company maintains that no serious harmful exploits have been demonstrated.
The move represents a significant escalation in U.S. AI export controls. Previous restrictions largely focused on advanced semiconductors and computing hardware, whereas this directive directly targets access to frontier AI models. Because limiting access by nationality is technically and operationally difficult, Anthropic opted to disable the affected models broadly in order to comply with the order.
The decision also comes amid ongoing tensions between Anthropic and parts of the U.S. government regarding military and national-security uses of AI. Anthropic has previously resisted allowing its technology to be used for domestic surveillance and fully autonomous weapons, leading to disputes with defense officials. The latest suspension highlights the growing intersection of AI development, cybersecurity, export controls, and geopolitical competition as governments seek greater oversight of the world's most powerful AI systems.
