According to reports from the Wall Street Journal, Amazon's cybersecurity research played a pivotal role in the White House's decision to cut off access to Anthropic's Fable 5 and Mythos 5 models. The research suggested that through specific prompts, these AI systems could provide information useful for cyberattacks.

Notably, CEO Andy Jassy met with officials from the White House prior to this directive, sharing Amazon’s findings on security risks. Following this exchange, the call was made to restrict access by foreign nationals, a decision that complicates matters given that many of Anthropic's researchers are themselves foreign-born.

In response, Anthropic disputed the government’s characterization, arguing that similar vulnerabilities could be found in other models too. Critics suggest there may have been underlying political motivations at play; former Commerce Department official Kate Koren hinted that the White House’s stance on Anthropic might have influenced the decision.

The long-standing tensions between Anthropic and the Trump administration over AI ethics resurface, with the company refusing to allow its technology for mass surveillance or lethal autonomous weapons. Despite temporary amends, the latest development signals a potential clash again.