Anthropic is pulling its two newest AI models after a U.S. government directive

Anthropic, the prominent AI developer, has abruptly disabled its two newest and most advanced artificial intelligence models, Claude Fable 5 and Mythos 5, for all global users following an unprecedented export control directive from the U.S. government. Issued by Commerce Secretary Howard Lutnick on Friday, June 12, 2026, just three days after their launch, the order cites unspecified national security concerns, marking a significant escalation in Washington's bid to control frontier AI technologies and raising immediate questions about future AI deployment. The directive, part of the Trump administration increasingly aggressive stance on AI, targets 'any foreign national' both within and outside the U.S., including Anthropic own overseas employees. Unable to reliably filter users in real time, Anthropic opted for a complete shutdown, disrupting service for potentially hundreds of millions of users, including India's substantial Claude user base. While the government alluded to a 'narrow, non-universal jailbreak' in Fable 5—a technique to identify software flaws—Anthropic vehemently disagrees, stating that comparable capabilities are routinely available in models like OpenAI's GPT-5.5 without similar restrictions. This action follows a contentious history between Anthropic and the Trump administration, including a previous 'supply chain risk' designation after Anthropic refused Pentagon deals involving surveillance and autonomous weapons. Anthropic, currently navigating an IPO sprint, has pledged to comply while openly disputing the rationale and promising to share further technical details within 24 hours. The company is working to restore access, but this incident sets a formidable precedent, effectively establishing export controls on specific AI models and signaling a new era of government intervention in AI development. The broader AI industry now faces profound uncertainty, with providers of frontier models contemplating the implications for future releases and the potential for similar directives to halt innovation.