U.S. Export Controls Force Anthropic to Suspend Two Latest AI Models

Anthropic has taken its two newest AI models, Fable 5 and Mythos 5, offline after receiving an export-control order from the U.S. government citing national-security concerns. The administration’s letter, sent on Friday, did not disclose specific risks but required the company to prevent foreign nationals, including many of its own employees, from accessing the models.

According to podcast participants, the order may have been triggered by reports that Amazon researchers discovered a method to bypass Fable 5’s guardrails, which Amazon CEO Andy Jassy reportedly raised with White House officials. The decision came quickly, prompting speculation that it could serve as a distraction from other diplomatic efforts.

Cybersecurity experts have signed an open letter urging the administration to revoke the order, arguing that removing advanced AI capabilities could hinder U.S. network defenders. They contend that the security risk posed by Anthropic’s models is not unique among leading AI labs.

Industry observers note that the action could have mixed implications for competitors. Some see a potential advantage for firms that avoid a similar crackdown, while others warn that the regulatory environment remains uncertain. Anthropic has defended its models, stating that similar jailbreak vulnerabilities exist in other AI systems.

The episode also discussed how the controversy might affect Anthropic’s market perception. Following the suspension, downloads of the company’s Claude assistant increased, suggesting heightened public interest in alternatives to dominant chatbots.