Anthropic announced it has suspended access to its artificial intelligence models Fable 5 and Mythos 5 in compliance with an export control directive from the US government.
The company reported that at 5:21 PM it received an order requiring the suspension of access to these models for all foreign nationals, regardless of whether they are inside or outside the US, including Anthropic’s own foreign employees.
Anthropic said it immediately cut off access for all customers to fully comply with the directive. The company also emphasized that other AI models will not be affected by this decision.
Decision Comes Days After Launch of New Models
The directive came just days after Anthropic announced Fable 5 and Mythos 5, presenting them as high-performing systems across various industry benchmarks.
Fable 5 attracted particular attention as it was the first time the company publicly released such an advanced model. Anthropic noted that this version includes new safety measures to block responses in certain high-risk areas.
The models are built on Claude Mythos Preview, which the company introduced in April. This preview version had garnered interest in Wall Street and government circles due to its advanced cybersecurity capabilities.
Anthropic stated it does not plan to release this model for general use and offers it only to select companies under its cybersecurity initiative Project Glasswing.
Company Says It Has Not Been Informed of the Reasoning Behind the Decision
In its Friday statement, Anthropic said the US government did not provide details on which national security concerns justified the directive. The company apologized to its customers for the disruption caused.
The statement argued that unsafe uses should be prevented through a transparent, fair, open, and technically grounded legal process. Anthropic said the directive does not meet these principles.
This development marks the latest tension between the company and the US administration. Anthropic had previously come under scrutiny following a public dispute with the Department of Defense.
After talks between the parties failed, the Department of Defense labeled Anthropic as a supply chain risk. This designation reportedly requires defense contractors to certify that they do not use Anthropic’s Claude models in military projects.
Anthropic filed a lawsuit against the Trump administration to overturn this classification, and the legal process is ongoing.
