Legion LegalTech Files Lawsuit Over Anthropic AI Export Controls
Legion LegalTech, a legal technology company, has filed a high-profile lawsuit against the US government, challenging the recent ban on foreign nationals accessing Anthropic’s AI models. This case has stirred significant debate in the legal and technology sectors, especially around AI export controls and their impact on global innovation.
The Controversy Over AI Export Controls
The dispute centers on a directive issued by the US government on June 12, which required Anthropic to disable access to its advanced Fable 5 and Mythos 5 AI models for all foreign nationals worldwide. Legion LegalTech, which utilizes Fable 5 for product development and employs Canadian engineers, argues that this directive disrupted its operations and denied critical tools to its workforce.
According to the complaint, the AI export controls imposed by the government exceeded statutory authority. Legion contends that there is no current export control classification that regulates access to hosted AI models or their generated outputs. The company points to the rescission of ECCN 4E091 in May 2025, noting that no replacement provision exists. Thus, the lawsuit asserts that the Commerce Department cannot enforce a control that no longer exists on hosted AI models or their inferential text outputs.
Misapplication of Legal Authority
Legion LegalTech’s lawsuit alleges that the government misapplied export control laws, specifically the Export Control Reform Act (ECRA) and the International Emergency Economic Powers Act (IEEPA). The complaint challenges the use of Section 4817(b)(1) of the ECRA and a “Is Informed” letter issued by the Bureau of Industry and Security (BIS) to Anthropic.
Legion argues that Section 4817(b)(1) should only be invoked after formal rulemaking and coordination with international partners, processes which were not followed. Furthermore, “Is Informed” letters are typically case-specific, targeting particular customers or military end-users. In contrast, the government’s directive applied to all foreign nationals, regardless of their nationality, location, or intended use. The lawsuit describes this as an overbroad, blanket restriction that uses a targeted legal tool to impose a worldwide ban.
IEEPA and the Berman Amendment
A significant part of Legion’s argument focuses on the IEEPA and the Berman Amendment. The Berman Amendment expressly prohibits the government from using IEEPA to regulate the exportation of “information or informational materials” in any format. Legion claims that outputs generated by Anthropic’s AI models—such as drafted text, legal analysis, and summaries—qualify as informational materials protected by this exemption. Accordingly, the lawsuit asserts that the government does not have the authority to restrict AI-generated outputs under current law, further undermining the legal basis for the AI export controls.
Allegations of Arbitrary Action and Executive Conflict
The complaint also alleges that the government’s actions were arbitrary and capricious. Legion points out that the cited risk—a narrow jailbreak allowing model review of software code—resulted in a sweeping, worldwide suspension, which is both overbroad and underinclusive. The restrictions applied equally to all foreign nationals, while similar AI capabilities remained available through competing offerings like OpenAI’s GPT-5.5. Legion further criticizes the government for not considering less-restrictive alternatives, such as limiting the ban to adversary nations.
Another key argument involves a June 2 executive order signed by President Donald Trump. The order explicitly stated that it should not be interpreted as authorizing mandatory government licensing or preclearance for developing or distributing new AI models. Legion contends that the June 12 directive directly contradicts this executive order, further demonstrating the lack of statutory power behind the government’s action.
Broader Implications for the Legal Tech Industry
The outcome of Legion LegalTech’s lawsuit could have far-reaching implications for AI export controls and the use of hosted AI models by multinational teams. Legal technology companies that rely on global talent and advanced AI tools are closely watching this case, as it may set important precedents regarding access, compliance, and innovation in the sector.
Conclusion
The clash between Legion LegalTech and the US government highlights the ongoing challenges in regulating cutting-edge technologies like AI. As legal frameworks struggle to keep pace with innovation, the debate over AI export controls will likely intensify. The final outcome of this legal battle could shape future policies and determine how law firms and technology companies navigate the complex intersection of technology, law, and international collaboration.
This article is inspired by content from Original Source. It has been rephrased for originality. Images are credited to the original source.