The Trump administration imposed export controls on Anthropic’s Fable 5 AI model following urgent warnings from Amazon CEO Andy Jassy. Researchers at Amazon reportedly discovered the model could be manipulated to facilitate cyberattacks, prompting immediate federal action.
White House officials acted reluctantly after Anthropic CEO Dario Amodei resisted voluntary safety fixes. David Sacks of the President’s Council of Advisors on Science and Technology stated the administration was surprised by the lack of cooperation regarding the security vulnerability. Anthropic maintains the directive stems from a misunderstanding of a non-universal jailbreak and is working to restore access.
This unprecedented federal intervention demonstrated the government's ability to instantly restrict domestic AI infrastructure, triggering a capital rotation into decentralized alternatives. Bittensor surged nearly 24%, while Venice Token and Near Protocol also posted significant gains as investors sought censorship-resistant AI protocols.
Officials indicate export controls will remain until Anthropic remediates the identified safety issues. The incident underscores growing tension between centralized AI developers and national security mandates, accelerating market interest in decentralized AI infrastructure.