Amazon CEO reportedly raised Anthropic model concerns before government crackdown
Amazon CEO Andy Jassy reportedly alerted officials to security risks in Anthropic models, leading to government-imposed export restrictions on two AI systems.Summary
Amazon CEO Andy Jassy reportedly informed U.S. government officials that Amazon researchers discovered vulnerabilities in Anthropic's Claude Fable 5 model that could facilitate cyberattacks. This communication preceded a government-imposed export control ban on Anthropic's Fable 5 and Mythos 5 models. While Amazon declined to discuss specific private consultations, reports indicate the company raised these security concerns prior to the regulatory crackdown. David Sacks, co-chair of the President’s Council of Advisors on Science and Technology, noted that the administration requested Anthropic CEO Dario Amodei address the identified 'jailbreak,' but Amodei reportedly refused, resulting in the subsequent service restrictions.
(Source：TechCrunch)