Amazon chief executive Andy Jassy raised alarms about the security of Anthropic’s AI models during a meeting with Treasury Secretary Scott Bessent and other U.S. officials, according to a Wall Street Journal report. Jassy told the officials that Amazon researchers had used Anthropic’s Claude Fable 5 to extract information that could potentially be leveraged in cyberattacks. The warning set off a chain reaction that culminated in the Commerce Department imposing export controls on both the Fable 5 and Anthropic’s Mythos 5 models.
Government officials acted quickly. Within days, the export‑control ban went into effect, effectively cutting off worldwide access to the two models. The move marked one of the first high‑profile instances of the United States using export restrictions to curb the spread of advanced AI capabilities deemed risky.
Amazon’s role and response
Amazon, which has invested heavily in Anthropic, said it regularly engages with regulators on potential security threats but does not disclose the specifics of those conversations. A company spokesperson told reporters that it is "not uncommon for governments to seek our counsel on potential security risks" and reiterated that Amazon does not share the details of its discussions with officials.
The Information and Reuters echoed the Wall Street Journal’s account, reporting that Amazon had communicated concerns about the security of Anthropic’s models. The reports highlighted the company’s dual position as a cloud provider and a major investor, a combination that gives Amazon a unique perspective on the capabilities and risks of the models it hosts.
Government officials and Anthropic’s reaction
David Sacks, a former AI czar under the Trump administration and now co‑chair of the President’s Council of Advisors on Science and Technology, offered his own version of the exchange. He said a "highly credible trusted partner of both Anthropic and the USG" presented a jailbreak that could bypass the model’s safeguards. According to Sacks, the administration asked Anthropic CEO Dario Amodei to either fix the jailbreak or de‑deploy the model. Amodei allegedly refused.
Anthropic responded with a blog post acknowledging the concerns but arguing that the capabilities in question are already available in other publicly accessible models. The company updated the post to include a statement from an Amazon spokesperson, reinforcing the narrative that the issue is not unique to Anthropic’s offerings.
The episode underscores the growing tension between rapid AI development and national security considerations. While Amazon insists it is merely providing counsel, the export‑control action signals a willingness by the U.S. government to intervene when AI tools cross a perceived risk threshold.
Industry observers note that the ban could have ripple effects for other AI firms that rely on cloud infrastructure. With Amazon Web Services (AWS) reportedly affected by the model cut‑off, Anthropic may need to adjust its deployment strategy or seek alternative hosting solutions.
As the debate continues, the incident illustrates how private‑sector insights are increasingly shaping public policy on emerging technologies. Whether the export controls will prompt broader regulatory frameworks remains an open question, but the immediate impact on Anthropic’s model availability is already evident.
Este artigo foi escrito com a assistência de IA.
News Factory APP - notícias agênticas para impulsionar seu SEO e AEO.