The Trump administration shut down Anthropic’s Fable 5 and Mythos 5 models for foreign nationals after export controls tied to security concerns. Reporting described a vulnerability discovered through prompts used by Amazon researchers, where the models were induced to generate code patches that could enable exploitation if misused. Multiple accounts tie the trigger to cybersecurity risk testing and subsequent calls between Anthropic CEO Dario Amodei and senior administration officials. Anthropic said it prohibits access from within China and disputed aspects of how the threat was raised, while U.S. officials relied on national-security export authority to restrict distribution. For higher education and research, the immediate impact is access and collaboration risk: when leading AI systems are constrained by export controls, labs and researchers working on frontier tooling may face sudden limitations, compliance burdens, and workflow redesigns.