Startup Sues US Gov for Arbitrarily Killing Anthropic's Claude Fable 5

The San Jose startup Arthur Rothrock founded, Legion LegalTech, filed a lawsuit in a Washington federal court, demanding a judge strike down the June 12 directive that forced Anthropic to kill Fable 5 and Mythos 5. The startup builds AI tools for lawyers to draft lawsuits and filing documents, meaning the sudden death of their core engine was a literal existential threat.

The lawsuit points out that the government's export control powers are a total mess because the specific classification covering advanced AI weights was quietly axed in May 2025 and never replaced. Besides, sending text back and forth to an AI isn't "exporting technology" since users never actually get their hands on the actual code or weights.

The legal team also claims the ban is completely arbitrary, especially since OpenAI's GPT-5.5 is allowed to run the exact same code-analysis features that supposedly made Anthropic a national security threat. The lawsuit alleges the government made this massive decision in a single evening after a few cozy calls from Anthropic's rivals.

Adding to the comedy of errors, the directive was issued just ten days after a presidential executive order explicitly promised there would be no mandatory government licensing or pre-approval for AI models. To top it off, Anthropic itself is refusing to join the fight, choosing instead to play nice and issue diplomatic thank-you notes to the government while their own customers do the heavy lifting in court.

If this arbitrary ban stands, it sets a terrifying precedent where any bureaucrat can turn off a company's entire tech stack on a whim. While the developers of these models try to play nice with regulators to secure their monopolies, it is the small businesses relying on them that are left to bleed out in the name of fake national security.

Source: Yahoo News