Amazon Raised AI Security Concerns Over Anthropic Models
Laptop displaying “ACCESS DENIED” with documents on a desk
Key Highlights
Amazon CEO Andy Jassy reportedly alerted the Trump administration to security risks in Anthropic’s AI models, leading to a global shutdown after a bypass was discovered.
Amazon CEO Raised AI Security Concerns
Amazon CEO Andy Jassy reportedly raised concerns to senior Trump administration officials about security risks in Anthropic’s most advanced AI models. This intervention is believed to have contributed to Anthropic’s decision to globally shut down its latest models, Fable 5 and Mythos 5, on Friday.
US Government Mandates Shutdown
The US government informed Anthropic that a method existed to bypass safeguards, allowing the models to identify cybersecurity vulnerabilities. Although Anthropic stated the bypass only found “minor” flaws, the Trump administration issued an export control.
- The order mandated Anthropic to block any foreign nationals, both inside and outside the US, from accessing these models.
- Consequently, Anthropic disabled global access to Fable 5 and Mythos 5.
An Amazon spokesperson confirmed that governments often seek their counsel on potential security risks but declined to share specific discussion details.
Export Control Criticized
White House adviser David Sacks indicated the export control was issued reluctantly after Anthropic CEO Dario Amodei allegedly refused to fix the jailbreak or de-deploy the model. Some experts criticized the Trump administration’s action as poorly thought out.
The broad impact of the export control affects allied nations, including Canada and the UK, by preventing their nationals employed at Anthropic from conducting research and development.