Beyond Data Sovereignty: The AI Dependency Your Continuity Plan Doesn't Cover

On 12 June 2026, Anthropic disclosed that a U.S. export-control directive forced it to suspend access to two of its frontier models (Fable 5 and Mythos 5) for foreign nationals, and that complying meant switching the models off for affected customers. No breach. No outage. No contract violation. A fully operational, fully paid-for system was withdrawn overnight by a rule its users had no part in writing. Set aside the specifics; the event made concrete what most AI governance frameworks still treat as theoretical.

The systemic impact lands hardest on an enterprise that has embedded this model into a load-bearing process.