Amazon CEO Andy Jassy Triggered Government Crackdown on Anthropic's Fable AI Models

Amazon CEO Andy Jassy and executives from five other tech companies warned the Trump administration about security vulnerabilities in Anthropic's Fable model, triggering a government crackdown that forced the AI lab to shut down its most advanced models within hours. The White House issued an export control order on Thursday evening, giving Anthropic just 90 minutes to comply before blocking access to Fable 5 and Mythos 5 worldwide.

According to Axios, the government's move came after Amazon handed officials a report claiming that parts of the Fable model could be unlocked through jailbreaking, potentially aiding cyberattacks. The report was delivered on Thursday evening, and by 10:00 p.m. ET, Anthropic had shut down the models.

Investor Turns on Portfolio Company

Amazon is one of Anthropic's largest investors, making the company's role in the crackdown unusual. Anthropic trains and runs its models on Amazon's own AI chips. An Amazon spokesperson told Axios that it is not uncommon for governments to seek the company's counsel on potential security risks. At least five other companies also raised alarms with senior officials on Thursday evening and Friday morning.

• National Cyber Director Sean Cairncross called a meeting with top White House officials after the reports.
• The government tried for hours to convince Anthropic to pull the model voluntarily, but the company refused.
• Cybersecurity expert Katie Moussouris, who reviewed the Amazon report at Anthropic's request, called the response disproportionate. She said the technique flagged was "Defense Oriented Prompting" (DOP), used by defenders, not attackers.
• The export restriction is unlikely to be extended to other AI labs, according to The Information.

Government Message or Security Threat?

A source told Axios that the escalation may have been less about the actual security risk and more about the broader relationship between Anthropic and the U.S. government. Officials were bothered by what they saw as a lack of seriousness in how Anthropic handled the Fable release. The export restriction only came after Anthropic ignored an earlier request to pull the model on its own. "Companies will not screw with the White House. That is the ultimate effect," a person familiar with the situation told Axios, calling it a de facto licensing regime.

Anthropic had publicly welcomed an earlier executive order calling for a security review but released Fable before the review process was in place. The incident signals the U.S. government asserting state control over AI technology and restricting other countries' access. What comes next remains unclear, but the crackdown has set a precedent for how the government may handle future AI model releases.