Trump lifts restrictions on Anthropic models Mythos and Fable

The US Trump administration has lifted restrictions on Anthropic's AI models Mythos and Fable. The move follows a period of inconsistent AI policy from Washington, leaving companies in the sector with little clarity on the rules that will govern future model releases.

Anthropic, founded in 2021 by former OpenAI employees, focuses on safe AI development. The company had Mythos and Fable subject to federal restrictions whose precise content and legal basis have not been fully disclosed. Now that those restrictions have lapsed, the models can in principle be deployed or rolled out more broadly.

According to TechCrunch, the Trump administration's approach leaves companies like Anthropic in a state of uncertainty: they no longer know which rules apply when they want to bring new models to market.

What is known about the models and the restrictions

Mythos and Fable are AI models from Anthropic, the company behind the Claude assistant. Little is publicly known about the precise nature of the restrictions that were previously imposed. The source material provides no definitive answer as to whether they involved export controls, safety requirements, or other regulatory measures.

What is clear is that the decision to lift the restrictions fits within a broader pattern of shifting policy choices by the Trump administration in the field of artificial intelligence. The administration had earlier revoked AI safety executive orders introduced under President Biden. Since then, companies like Anthropic have been without a clear picture of which rules govern the development and deployment of new models.

Anthropic has received billions of dollars in investment in recent years from, among others, Google and Amazon, and explicitly positions itself as a lab that places safety at the centre of large language model development. That positioning makes the company particularly sensitive to uncertainty around federal safety frameworks.

Erratic AI policy as a constant factor

The decision on Mythos and Fable does not stand alone. Since taking office in January 2025, the Trump administration has repeatedly shifted course on technology regulation and AI policy, leaving companies without a stable legal framework to guide product development and market introduction.

For AI companies, that lack of clarity has practical consequences. Decisions about when a model is ready for commercial deployment, which safety checks must be completed, and how to communicate with federal regulators all become more difficult when the policy landscape changes rapidly and unpredictably.

TechCrunch describes the administration's approach as erratic, a characterisation that is widely recognised across the industry. Several major AI labs and trade associations have in recent months called for greater consistency in federal policy, though that has so far produced little visible result.

Implications for the broader AI landscape

For European and Dutch players in the AI sector, the situation in the United States is relevant as a point of reference. The EU is working through the AI Act to establish a structured regulatory framework that aims to provide precisely the long-term certainty currently absent in the US. That contrast could increase Europe's attractiveness as a location for AI development, although much depends on how swiftly and consistently the AI Act is applied in practice.

For investors and founders operating in markets with transatlantic ambitions, the events surrounding Mythos and Fable highlight how strongly regulatory uncertainty can affect planning around model releases. Anyone dependent on access to the US market must account for policy shifts that are difficult to anticipate based on prior signals.