Amazon CEO & US Officials: Anthropic AI Ban

Conversations between Amazon CEO Andy Jassy and senior US government officials influenced the decision to block foreign access to Anthropic's advanced AI models, Fable 5 and Mythos 5. Here's the thing: Jassy told officials, including US Treasury Secretary Scott Bessent, that Amazon researchers bypassed Fable 5's safety protections. They used a series of prompts to get the model to provide information that could potentially aid cyberattacks. This information was "supposed to be off limits." What's interesting is that the US government then banned foreign access to both Fable 5 and Mythos 5. This means foreign governments, companies, and nationals can no longer use these models. Anthropic had to restrict access entirely due to the broad scope of the rule. The company stated the government did not provide specific details about its national security concerns. Anthropic also said the "jailbreak method" cited by officials exposed only a small number of previously known and minor vulnerabilities. They argue other publicly available AI models could identify similar vulnerabilities without bypassing safety guardrails. The bottom line is this highlights ongoing concerns about AI models and their potential national security risks.