Amazon CEO Allegedly Voiced Worries About Anthropic Model Prior to Government Intervention
Concerns raised by Amazon’s CEO, Andy Jassy, have resulted in Anthropic restricting global access to two of its models.
According to reports, Jassy informed Treasury Secretary Scott Bessent and other officials that researchers at Amazon might have used Anthropic’s Claude Fable 5 to obtain potentially harmful information for cyberattacks. In response, the government enforced an export control ban on the Fable 5 and Mythos 5 models.
An Amazon representative stated that while it is common for governments to consult the company regarding security risks, they do not reveal specifics of those discussions.
The spokesperson highlighted that Amazon Web Services (AWS) has been impacted by this decision regarding model access.
Other sources have echoed reports that Amazon, a significant investor in Anthropic, expressed concerns regarding the security of the company’s models.
David Sacks, who previously held a position as AI czar under Trump and currently co-chairs the President’s Council of Advisors on Science and Technology, shared his perspective on the situation, claiming that a trusted partner associated with both Anthropic and the government raised an issue regarding a potential security flaw.
Sacks noted that the administration requested that Anthropic’s CEO, Dario Amodei, address the security risk or withdraw the model entirely, but Amodei declined to do so.
In a blog update, Anthropic asserted that the features prompting governmental concern are already present in other widely available models.
This article has been updated with a statement from an Amazon spokesperson.
