Anthropic's 'Fable' Model Just Vanished Under New Export Controls — and the Real Reason Is Stranger Than You Think

The Trump administration last Friday imposed export controls on Anthropic's latest AI model, codenamed "Fable," restricting access to U.S. citizens. Anthropic, no longer able to legally distribute the model domestically, made it unavailable entirely. The move has sparked debate over the administration's rationale. As Ben Thompson notes in Stratechery, Occam's Razor suggests that policymakers still don't fully understand how AI works, leading to heavy-handed regulation.

Anthropic faces a unique dilemma: its safety-first reputation gives it license to aggressively prioritize its business and even challenge the U.S. government, yet every action it takes appears from the outside to be self-serving. This incident underscores the friction between AI safety advocacy and national security interests. The broader implications include a chilling effect on model releases and a potential shift in how frontier labs navigate export controls, especially as the line between safety and censorship blurs.