Key Facts

• Anthropic’s advanced AI models, Claude Fable 5 and Mythos 5, were restricted from global access by the US government on June 12.
• The restrictions were lifted after the company addressed national security risks in close cooperation with the US government.
• The Department of Commerce officially withdrew the export controls and allowed Anthropic to restore access starting the day after the announcement.
• Earlier, only a small group of US cybersecurity firms could access one model, Mythos 5, under government authorization.
• The Trump administration led these controls and praised companies for working together to improve AI security.
• OpenAI, a competitor, also limited access to its new model GPT-5.6 following US government guidance.
• The administration issued an executive order in early June requiring steps for AI safety and cybersecurity involving private tech firms.
• CIA Director John Ratcliffe compared top AI models’ power to nuclear weapons, highlighting government concerns about their risks.