The White House administration imposed export restrictions on Anthropic models, with the key trigger being concerns about possible access to the Mythos model by a group linked to China. This incident, along with reports of attempts to bypass the protections of the Fable 5 model, formed the basis for the authorities' decision.
Double Trigger: China and Security Bypass
According to available information, suspicions of access to Mythos by a Chinese group were one of the main factors that prompted the Donald Trump administration to take action. However, the exact organization, the method of gaining access, and the source of information for the White House remain undisclosed. The second trigger was data indicating that Amazon CEO Andy Jassy informed the administration on June 11 about the possibility of bypassing Fable 5's restrictions, in response to a request for model review.
U.S. Secretary of Commerce Howard Lutnick sent a letter to Anthropic CEO Dario Amodei, stating that Fable 5 and Mythos 5 were subject to export controls outside the U.S. and for all foreign individuals within the country. Officials attempted to persuade the company to voluntarily suspend access, but Anthropic refused. According to a source close to the startup, the White House gave the company only 90 minutes to disable the models, without providing details about the nature of the threat.
White House advisor David Sacks stated that the administration received a warning from a partner trusted by both the authorities and Anthropic. He noted that "the ball is in Anthropic's court" and that the export controls would be lifted once the security issue is resolved.
Anthropic's Position: "Misunderstanding"
Anthropic categorically denies the allegations. The company claims that the authorities' letter contained no specific details about national security risks, and the complaints only concerned the method of bypassing Fable 5's protections. The startup examined a demonstration of this technique and concluded that it could only find "a small number of previously known minor vulnerabilities." Moreover, other publicly available models can identify the same issues without bypassing protections.
Before launching Fable 5, the company spent "thousands of hours" testing safeguards jointly with the U.S. government, the UK AI Safety Institute, and third-party organizations. Testers found no universal way to massively bypass the model's restrictions. Anthropic considers the incident a "misunderstanding" and is working to restore access.
As a reminder, on June 9, Anthropic introduced Fable 5 — a Mythos-class model for general access, and Mythos 5 — a model with relaxed restrictions for cybersecurity professionals under Project Glasswing. Earlier, on June 2, the company announced the expansion of the project to approximately 150 organizations from over 15 countries, which have already identified more than 10,000 high and critical severity vulnerabilities.
Expert Opinion: The situation around Anthropic is a classic example of how political risks and national security concerns begin to dominate technological progress. Even if the allegations of "Chinese access" are exaggerated, the very fact of such a harsh reaction from the authorities signals a new era of AI regulation, where export controls will become a tool of geopolitical pressure. Investors should consider that such incidents will increase volatility in the sector, as we have already seen with the rise of Venice and Morpheus tokens by 14% and 21%, respectively, following the imposition of restrictions.