A legal tech firm has filed a lawsuit against the US government, challenging an order that limits foreign access to Anthropic’s top-tier AI models. The firm claims the directive violates existing export control laws and has caused immediate harm to its operations. The lawsuit was filed in a federal court in Washington, DC, and seeks to vacate the administration’s directive.

Legion LegalTech, based in San Jose, California, relies on Anthropic’s tools for its software platform. The US government’s action, which came in the form of a directive from the Trump administration, cut off access for its Canadian-based users. The company argues that the harm caused by the order is immediate, irreparable, and existential.

The lawsuit specifically targets a directive issued on June 12, which ordered Anthropic to disable its Fable 5 models. Legion claims that the US government’s use of export-control laws to restrict access to hosted AI models or their outputs is not supported by existing legal frameworks. The case is ongoing, with no immediate resolution in sight.

The legal battle raises broader questions about the regulation of AI models and the potential for export controls to impact global access. The outcome could influence how AI companies operate internationally and how governments enforce export laws. The case also highlights the growing tension between AI innovation and regulatory oversight.

As the case continues, the legal and regulatory landscape for AI remains uncertain. The outcome may set a precedent for future disputes involving AI export controls and the rights of foreign users. The litigation underscores the complex interplay between technology, law, and international policy.