Amazon CEO Flags Anthropic AI Security Concerns

Anthropic's newest AI models, Fable 5 and Mythos 5, have been pulled offline globally due to security concerns. Reports suggest Amazon CEO Andy Jassy flagged these concerns to U.S. officials. Jassy reportedly told Treasury Secretary Scott Bessent that Amazon researchers used Anthropic's Fable 5 model to obtain information potentially usable in cyberattacks. Shortly after, the government restricted access to both models. Anthropic strongly disputes this characterization. This situation is unusual because Amazon is a major investor in Anthropic, committing billions of dollars. The dispute also follows Anthropic CEO Dario Amodei's public warning about rapid AI development and the need for stronger safeguards. Government officials state Anthropic declined to fix or remove the models after researchers bypassed their guardrails, a claim Anthropic contests. This could impact Anthropic's anticipated IPO.