Anthropic disables Fable and Mythos AI models following U.S. government export ban

Key Points:

• The U.S. government issued a directive requiring Anthropic to disable its powerful AI model Fable 5 for all users, including foreign nationals within the U.S., due to national security concerns related to a potential jailbreak that could unlock cybersecurity capabilities.
• Anthropic argued the jailbreak was narrow and not a universal threat, contending that similar vulnerabilities exist in other AI models like OpenAI’s GPT-5.5, which are not subject to such export controls, and criticized the government’s action as lacking transparency and fairness.
• This directive follows previous tensions with the Trump administration, which banned federal use of Anthropic’s models and labeled the company a “supply chain risk,” leading to ongoing legal challenges and potential impacts on Anthropic’s planned IPO and investor confidence.
• Industry experts and policy analysts expressed disbelief and concern over the directive, highlighting inconsistencies in U.S. AI policy, potential harm to innovation, and the risk of driving Chinese-born AI researchers abroad.
• Critics also noted that Anthropic’s own framing of its AI models as potentially dangerous may have contributed to government scrutiny, while historical attempts to restrict software exports have largely failed.