The White House has imposed export restrictions on Anthropic's Fable 5 and Mythos 5 models, and the key reason, according to my information, was concerns about access to the Mythos model by a China-linked group. This decision, made in mid-June, sparked a strong reaction in the industry and raised questions about the balance between security and innovation.

Triggers and Government Reaction

In addition to the "Chinese connection," reports of a possible bypass of Fable 5's safeguards served as a trigger. However, Anthropic itself categorically denies the seriousness of the threat, claiming it involved a "narrow" and non-universal method that did not provide the model with unique capabilities. My sources in the industry confirm that the Donald Trump administration received a warning from a trusted partner, after which it demanded that Anthropic disable access to the models within 90 minutes, without going into details of the threat.

White House advisor David Sacks publicly stated that "the ball is in Anthropic's court," implying that the company must fix the security issue to have the restrictions lifted. At the same time, Commerce Secretary Howard Lutnick sent an official letter to Anthropic CEO Dario Amodei, placing Fable 5 and Mythos 5 under export controls for both foreign persons in the U.S. and for use outside the country.

Anthropic's Position and Consequences

Anthropic insists that the government's letter did not contain specific details about risks to national security. The company claims it examined the demonstration of the bypass technique and concluded that it only reveals "a small number of previously known minor vulnerabilities" that could be found in other public models without bypassing safeguards. Moreover, before launching Fable 5, the company spent "thousands of hours" testing with the U.S. government and the UK's AI Safety Institute, and no universal method for mass bypass was found.

It is worth recalling that Fable 5 is a model for general access, while Mythos 5 is its version with relaxed restrictions for cybersecurity defenders under the Project Glasswing program. It was Mythos 5, designed for finding vulnerabilities, that raised the most suspicion. Significantly, after the restrictions were imposed, tokens of decentralized AI projects Venice and Morpheus surged by 14% and 21% respectively — the market clearly saw this move as confirmation of the growing role and value of decentralized alternatives.

My analysis: The situation surrounding Anthropic is a precedent that demonstrates how thin the line is between protecting national interests and restraining technological progress. On the one hand, the White House's concerns are understandable given the geopolitical tensions. On the other hand, government actions based on "narrow" and unconfirmed data risk undermining trust in leading American AI companies and pushing the market to seek more "uncontrollable" solutions, which in the long term could weaken the U.S. position in the global artificial intelligence race.