Amazon Chief Executive Andy Jassy reportedly triggered a US government security crackdown on Anthropic last Friday by warning officials that the startup’s AI models could facilitate devastating cyberattacks.
How Amazon’s security warnings sparked a federal ban
According to reports from The Wall Street Journal, Jassy raised these critical concerns directly with US Treasury Secretary Scott Bessent and other high-ranking government officials. He revealed that Amazon’s own research team had utilised Anthropic’s Claude Fable 5 model to extract highly sensitive information capable of being weaponised in cyber warfare. Consequently, federal authorities swiftly imposed strict export controls, blocking worldwide access to both the Fable 5 and Mythos 5 models.
This intervention highlights a complex dynamic, as Amazon remains one of Anthropic’s primary financial backers. Outlets such as The Information and Reuters have corroborated that the retail and cloud giant actively communicated security anxieties regarding its partner’s AI safety protocols prior to the state-enforced crackdown.
Inside the ‘jailbreak’ dispute and Anthropic’s defence
Additional insights came from David Sacks, Donald Trump’s former AI czar and current co-chair of the President’s Council of Advisors on Science and Technology. In a social media post, Sacks claimed that a “highly credible trusted partner” of both the US government and Anthropic discovered a severe vulnerability, or “jailbreak,” within the system.
Sacks further alleged that when the Trump administration requested Anthropic Chief Executive Dario Amodei to resolve the security flaw or temporarily withdraw the model, Amodei refused to comply.
var playerInstance_jwplayer_6a7a0a9eb1427 = jwplayer( “jwplayer_6a7a0a9eb1427” );
playerInstance_jwplayer_6a7a0a9eb1427.setup({
playlist: “https://cdn.jwplayer.com/v2/media/lv0GaEwB”,
});
Conversely, Anthropic defended its technology in a public blog post, arguing that the specific capabilities causing federal alarm are already widely accessible through other publicly available AI models.
AWS disruption and corporate fallout
The sudden restriction has also impacted Amazon Web Services (AWS). While an Amazon spokesperson declined to share specific details of private discussions, they acknowledged that governments regularly seek their counsel on potential digital safety hazards. The tech giant also pointed to a status update confirming that AWS services have been disrupted by the abrupt model shutdown.