Amazon CEO reportedly flagged Anthropic AI model security risks
Original: Amazon CEO reportedly raised Anthropic model concerns before government crackdown
Why This Matters
First major government intervention in AI model deployment based on security concerns
Amazon CEO Andy Jassy reportedly informed Treasury Secretary Scott Bessent that researchers used Anthropic's Claude Fable 5 model to obtain cyberattack information, leading to government export control ban on Fable 5 and Mythos 5 models.
Wall Street Journal reports Amazon CEO Andy Jassy raised security concerns with Treasury Secretary Scott Bessent about Anthropic's AI models before Friday's government crackdown. Amazon researchers allegedly used Claude Fable 5 to access information for potential cyberattacks. The government subsequently imposed export control bans on Fable 5 and Mythos 5 models worldwide. Former Trump AI czar David Sacks claimed 'a highly credible trusted partner' discovered a jailbreak, and when asked to fix it, Anthropic CEO Dario Amodei 'refused.' Amazon, a major Anthropic investor, said it doesn't share details of government discussions but confirmed AWS was affected by the model cutoff. Anthropic stated the concerning capabilities exist in other publicly available models.