THE REAL STORY OF THE FABLE 5 BAN: AMAZON CEO CALLED THE TREASURY, NOT THE PENTAGON, AND EVERYTHING SPIRALED The 20-day suspension of Claude Fable 5 that rattled enterprise AI customers worldwide was not a carefully planned regulatory action. It was a chain of panicked phone calls that escalated out of control. Axios has published the most detailed account yet of what actually happened, and the picture it paints is of the most improvised AI regulatory event in American history. The chain started with Amazon CEO Andy Jassy, who called Treasury Secretary Scott Bessent after Amazon researchers found a vulnerability in Fable 5 related to cybersecurity exploits. Bessent escalated to White House Chief of Staff Susie Wiles. Within hours, Commerce Secretary Howard Lutnick, directed by President Trump, called Anthropic CEO Dario Amodei directly to announce the export control order on June 12. What followed was 20 days of engineering work in Washington. Anthropic flew teams of engineers to DC. The NSA reviewed their initial fixes and rejected them as insufficient, forcing further rounds of changes. The final restoration came gradually, agency by agency, until all cleared on July 1. The most damaging detail is this: cybersecurity experts later confirmed in an open letter that other leading AI models, including GPT-5.5 and Kimi K2.7, contained the same vulnerability that triggered the Fable 5 ban. Those models faced no restrictions. The ban disrupted global enterprise AI for three weeks over a capability that already existed, untouched, in models nobody regulated. Keywords: Fable 5 ban story, Anthropic export control, Andy Jassy Scott Bessent, AI regulatory failure 2026