Anthropic disables access to its newly launched AI models Fable 5 and Mythos 5 after receiving a U.S. government directive tied to national security and export controls. Multiple outlets report that the order leads to Anthropic blocking access for customers, including potentially outside the United States, and that the company shuts down access in response to the government’s instructions. The government’s stated rationale centers on concerns that a “jailbreak” risk in Fable 5 could pose a national security threat, with some reports describing the directive as coming from the Commerce Department and referencing broader “national security authorities.” Anthropic acknowledges the government order but says its models present minimal jailbreaking risk and disputes the basis for the action. At least one outlet says Anthropic characterizes the cited vulnerabilities as known and minor. Several reports also note the timing: access is disabled shortly after the models’ launch, with references to a Friday action following the directive. Anthropic’s communications indicate it is complying with the order while continuing to push for clarification and the possibility of restored access.