AI developer Anthropic has taken its latest artificial intelligence models, Fable 5 and Mythos 5, offline following a directive from the Trump administration aimed at preventing their use by foreign nationals, marking the U.S. government’s most aggressive move to date in controlling access to advanced AI systems.
What Triggered the Removal of Fable 5 and Mythos 5
The export controls were imposed under a directive received by Anthropic on Friday afternoon from the U.S. government. Fable 5, a limited version of the more powerful Mythos 5, had only recently been released widely. Anthropic had already tightly controlled access to Mythos 5 due to cybersecurity concerns surrounding its advanced capabilities.
In a public statement, Anthropic said it disagrees with the government’s handling of the matter. The company noted that the directive did not specify the national security concerns behind the order. “We believe the government should have the ability to block unsafe deployments, as part of a statutory process that is transparent, fair, clear, and grounded in technical facts,” the company said. “This action does not adhere to those principles.”
Anthropic characterized the situation as a “misunderstanding” and expressed hope to restore access to the models “as soon as possible.” The Commerce Department did not immediately respond to a request for comment.
Executive Order Established the Framework
The action comes 10 days after President Donald Trump signed an executive order intended to create a framework for the federal government to vet the national security risks of the most advanced AI systems for up to a month before their public release. The order had described participation by AI developers as voluntary, making the directive to Anthropic a notable escalation in enforcement.
Why These Export Controls Matter for AI Security
What is the significance of the U.S. government restricting access to AI models like Fable 5 and Mythos 5? The move signals that the administration is willing to go beyond voluntary frameworks to enforce national security controls on advanced AI technologies. For the cybersecurity industry, this raises questions about how rapidly capabilities can be pulled back after release and what due process developers can expect. The models in question are reportedly capable of accelerating exploit creation — turning n-days into n-hours — according to prior analysis of Claude Mythos, making them a double-edged sword for both defenders and attackers.
Anthropic’s Disagreement With the Process
Anthropic’s public objection centers on the lack of transparency and technical grounding in the directive. The company has previously disputed claims about jailbreaking Fable 5, and the broader industry has watched closely as the balance between innovation and security control tightens. The absence of a clear statutory process, as Anthropic points out, creates uncertainty for any developer building powerful AI systems in the United States.
What Organizations Should Consider Now
Any organization developing or deploying advanced AI models should review their compliance posture regarding potential export control restrictions. Companies in this space should engage proactively with legal frameworks and prepare for the possibility of similar directives. For enterprises that rely on AI models from third-party developers, it is prudent to evaluate the continuity of access and the security implications of any sudden removal of capabilities. The incident underscores the need for a multi-layered approach to AI governance that accounts for both technical safeguards and regulatory compliance, without relying on any single vendor or model for critical operations.
