Anthropic Disables New AI Models After Government Warning

Anthropic has officially shut down its new AI models, Fable 5 and Mythos 5, following a directive from the United States government. The company received an order from the Commerce Department on Friday evening. This directive subjects the new models to strict export controls. Consequently, Anthropic has disabled access for all customers to ensure they comply with the law. The government’s decision is based on concerns about national security. According to reports, there is a risk of a jailbreak. A jailbreak occurs when users find a way to bypass the safety safeguards of the AI. If these safeguards fail, the model might provide information about biology, chemistry, or cybersecurity. An administration official suggested that the government needs more time to harden its security apparatus against these threats. This process might take a few weeks to complete. Anthropic has apologized for the sudden disruption, which they described as a misunderstanding. However, the company has expressed strong disagreement with the government's reasoning. They argued that the jailbreak is narrow and not universal. They also pointed out that other publicly available models, such as GPT-5.5, possess similar capabilities. Earlier this month, President Trump signed an executive order regarding AI safety. This order encourages companies to participate in voluntary security testing. Anthropic has promised to release more details about the situation within the next twenty-four hours. For now, only Fable 5 and Mythos 5 have been removed. Other models from Anthropic are still functioning normally. This event has caused a major conversation about how the government should regulate new technology and the balance between innovation and national safety.