Amazon warning triggered US crackdown on Anthropic AI models: Reports
The Trump administration’s decision to block foreign access to Anthropic’s most advanced AI models was reportedly prompted by warnings from Amazon CEO Andy Jassy and at least five other firms. Amazon researchers found a way to prompt Anthropic’s Fable 5 model into producing information that could aid cyberattacks, leading Jassy to contact senior US officials. The White House then assessed the risk and pressed Anthropic CEO Dario Amodei to voluntarily pull the model, but he reportedly resisted. David Sacks said the export control was issued reluctantly and that the government wants Anthropic to fix the jailbreak issue so access can be restored. Anthropic said the move stemmed from a misunderstanding about a “non-universal jailbreak” and said it is working to restore access. The episode also highlighted how quickly US regulators can cut access to domestic AI models, boosting decentralized AI-related tokens.
