Legal tech startup sues US government after losing access to Anthropic's Fable 5 AI model

Legal Tech Firm Sues US Government Over AI Access Restrictions

Legion LegalTech Corp has filed a lawsuit in Washington, D.C., federal court challenging a US government directive that resulted in Anthropic halting access to two of its most advanced AI models worldwide 1

1

. The Legion lawsuit targets a June 12 order from the Bureau of Industry and Security that required Anthropic to disable its Fable 5 and Mythos 5 models for any foreign national, both inside and outside the United States 2

2

. According to the complaint, the government gave Anthropic approximately 90 minutes to comply with the order or face civil and criminal penalties 4

4

. Anthropic responded by turning off access to the models for all customers the same day to ensure compliance 1

1

.

Immediate and Irreparable Harm to AI Innovation

The San Jose-based Legion LegalTech Corp says it depends on Anthropic AI models for its software platform, which builds drafting and case-management tools for attorneys 1

1

. The company's software can draft pleadings, discovery requests, motions, and other court documents . The US government directive immediately cut off access for members of Legion's Canada-based software development team and disrupted its business operations 1

1

. "The harm to Legion is immediate, irreparable, and existential," the lawsuit states. "The pace of frontier AI advancement is blistering, and competitive ground lost during a suspension cannot be regained after the fact" 1

1

. The company argues that each day the directive remains in force disrupts its product, sidelines its engineers, and erodes its ability to survive in a field defined by continuous access to the most capable models 3

3

.

Export Control Laws and Statutory Authority Challenged

The Legion lawsuit argues that the US government exceeded its statutory authority under export control laws and emergency powers legislation 5

5

. According to the filing, no current export control classifications cover access to cloud-hosted AI models or their outputs 4

4

. The complaint states that "the only export-control classification that ever directly covered advanced AI model weights -- ECCN 4E091 -- was rescinded in May 2025 with no replacement" and that "no currently operative provision of the Commerce Control List classifies access to a hosted AI model, or its inferential text output, as a controlled item" 5

5

. Legion also challenges the legal mechanisms the Commerce Department relied on, arguing that the Bureau of Industry and Security used "Is Informed" letters—normally used for specific cases involving identified military-intelligence concerns—as a mechanism for imposing blanket worldwide bans on foreign nationals regardless of nationality, location, or intended use 5

5

.

National Security Concerns vs. AI Development

Fable 5 and Mythos 5 were Anthropic's attempt to offer public versions of its more restricted Mythos-class technology . Fable 5 was released to the public on June 9 and taken down on June 12, making it available for just over three days 3

3

. The government's concern reportedly centered on claims that Fable 5 could be jailbroken, allowing users to bypass some restrictions . According to media reports, Amazon researchers had found workarounds that bypassed the safeguards, which combined with previous worries that China-connected entities had access to Mythos, apparently led to the export control order restricting foreign access to AI 3

3

. However, Anthropic stated it reviewed the technique and believed the demonstrated vulnerabilities were minor, already known, and discoverable by other public models . The company criticized the government for "not providing specific details of its national security concern" 4

4

.

International Emergency Economic Powers Act Violations Alleged

Legion's complaint further argues that if the directive relies on the International Emergency Economic Powers Act, it violates the Berman Amendment, which excludes from IEEPA authority any power to regulate "the exportation of information or informational materials in any format or medium of transmission" 5

5

. The lawsuit contends that AI-generated outputs, including drafted text, written legal analysis, and summaries, qualify as informational materials protected by this exemption 5

5

. Additionally, the filing notes that no national emergency has been declared with respect to the purported threat posed by the models, and the alleged threat did not have its source in whole or substantial part outside the United States as required under the statute 5

5

. The lawsuit also points to a June 2 executive order signed by President Donald Trump stating that nothing should be interpreted as authorizing "a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models," arguing that the June 12 directive imposed precisely such a requirement 5

5

.

What Comes Next for Legion and Anthropic

Legion asked a judge to vacate and set aside the US government directive targeting Anthropic and plans to seek a preliminary order to bar the administration from enforcing it 1

1

. The lawsuit seeks a declaration that the directive is unlawful, an order setting it aside, and an injunction preventing its enforcement 5

5

. Anthropic is not a party to the litigation but referred to a prior statement saying it was "grateful to the administration for their ongoing partnership in working to get this matter resolved as quickly as possible" 1

1

. The Commerce Department and White House did not immediately respond to requests for comment 1

1

. Anthropic and the United States are locked in separate legal battles in Washington and California federal courts, with Anthropic having sued the Trump administration after the government moved to place the company on a supply-chain blacklist over its refusal to allow the military to use its AI models for domestic surveillance or fully autonomous weapons 1

1

.