Anthropic suspends Fable 5 and Mythos 5 models following US export controls
The Trump administration’s order forces immediate compliance, prompting Anthropic to disable its latest models globally despite disputing the necessity of the recall.
Anthropic has abruptly suspended access to its Fable 5 and Mythos 5 artificial intelligence models following a directive from the US Commerce Department. The shutdown, which took effect late on Friday night, imposes strict export controls that restrict the use of these models to within the United States. The company stated that immediate compliance with the government order required disabling the models for all customers worldwide, leaving other Anthropic services unaffected.
The US government cited national security concerns regarding a reported 'jailbreak' that reportedly bypasses safety safeguards designed to block prompts related to cybersecurity, chemistry, and biology. According to an Axios report citing an administration official, the Trump administration requested a pause in the release of these models to allow time for the national security apparatus to be hardened against such threats, a process potentially expected to take a few weeks.
In its announcement, Anthropic noted that the government provided only verbal evidence of a potential narrow, non-universal jailbreak. The company explained that the specific concern involves getting Fable 5 to review a particular codebase for software flaws. Anthropic argued that the vulnerabilities identified are minor and relatively simple, and pointed out that similar capabilities exist in other publicly available models, such as GPT-5.5.
Despite complying with the directive, Anthropic disputed the necessity of the recall. The company warned that applying this standard across the industry would effectively halt all new model deployments for frontier providers. Anthropic described the situation as a misunderstanding and apologised to customers for the disruption, promising further details within 24 hours.
The suspension occurs against a backdrop of evolving regulatory scrutiny. Earlier this month, President Trump signed an executive order urging AI model makers to submit to voluntary government security testing. This followed a postponed signing ceremony last month, which was delayed amid reported disagreements within the administration regarding the scope of such measures.
