On June 12, 2026, the Trump administration imposed export controls on Anthropic's latest frontier AI models, Fable 5...

On June 12, 2026, the Trump administration imposed export controls on Anthropic's latest frontier AI models, Fable 5 and Mythos 5, citing national security concerns over potential jailbreak vulnerabilities. The directive bars any foreign national from accessing the models, forcing Anthropic to disable access for all users, including its own employees and US-based customers. This unprecedented use of export controls against a hosted AI service has alarmed allies and triggered a scramble among G7 nations to reassess reliance on US AI capabilities (Tom's Hardware).

The move has immediate commercial consequences: JPMorgan Chase and Goldman Sachs have banned staff in Hong Kong from accessing Anthropic's models (FT), and enterprise customers worldwide are reconsidering their AI dependencies. Anthropic's CEO Dario Amodei is lobbying at the G7 summit for a US-led AI coalition, but the damage to trust is evident (CNBC). The ban also accelerates sovereign AI initiatives in Europe and Asia, as seen in the rise of Chinese models like GLM-5.2 (Reddit).

For Anthropic, this represents a severe setback: its most advanced revenue-generating products are offline, investor confidence is shaken (Bloomberg), and competitors like OpenAI and xAI may capture market share. The direction is unequivocally down, with material impact on revenue, customer trust, and competitive positioning.