US Government Directs Anthropic to Disable Fable 5 and Mythos 5 AI Models Globally
Anthropic has disabled global access to its AI models, Fable 5 and Mythos 5, in response to an order from the US government. The government cited alleged jailbreak risks as the reason for its directive. While complying, Anthropic has publicly pushed back, stating that the vulnerabilities are minor and are also present in competing models like GPT-5.5.
The US government has ordered Anthropic to cease global access to its artificial intelligence models, Fable 5 and Mythos 5. The directive was issued due to alleged jailbreak risks associated with these AI systems.
Anthropic has confirmed its compliance with the government's order. However, the company has publicly contested the claims, asserting that the vulnerabilities are minor. It also highlighted that similar issues exist in rival models, including GPT-5.5.
Anthropic had previously spent months emphasizing the cybersecurity risks inherent in its own Mythos class models. The company has expressed concern that this action could set a precedent, potentially impacting the deployment of all future frontier AI technologies.
According to The Decoder AI, Anthropic warns that this move could establish a precedent that halts all frontier deployments.
