A leading United States think tank has filed legal support for artificial intelligence company Anthropic in its lawsuit against the Trump administration, warning that the Pentagon’s decision to blacklist the firm over its safety guidelines sets a dangerous precedent for governments seeking to control how private technology companies build and deploy their products.

Anthropic filed two federal lawsuits on Monday, March 9, seeking to reverse a Pentagon decision designating the company a supply chain risk after it refused to remove restrictions on its AI model Claude that prevent its use for autonomous weapons and mass surveillance of American citizens. The designation, historically reserved for foreign adversaries, requires defence contractors to certify they are not using Claude in Pentagon-related work and follows a directive by President Donald Trump ordering all federal agencies to stop using Anthropic’s technology.

The dispute centres on two firm red lines that Anthropic established for Claude: that it would not be used for mass surveillance of US citizens, and that it would not be deployed for fully autonomous weapons systems without human oversight. The Pentagon insisted it needed the ability to use Claude for “all lawful purposes” and that a private company could not constrain military decision-making.

The Cato Institute, a Washington-based libertarian think tank, joined an amicus brief supporting Anthropic, with its lawyers arguing that the Pentagon’s response amounted to unconstitutional retaliation against protected speech. Thomas Berry, Director of Cato’s Robert A. Levy Center for Constitutional Studies, wrote that the Defence Department “punished Anthropic because Anthropic refused to change its algorithm and jettison its guidelines to produce outcomes of the Pentagon’s choosing.”

In an exclusive interview with NewsGhana, Jennifer Huddleston, Senior Fellow in Technology Policy at the Cato Institute, said the case carries consequences that extend well beyond the immediate legal dispute.

On the broader precedent, Huddleston warned that a ruling in favour of the Pentagon could open the door to similar pressure being applied to other AI companies. “We could see similar pressure exerted over AI companies to change the safeguards or programming of their models to suit government demands,” she said. “This could allow the government to more broadly dictate how models must operate.”

She drew a sharp distinction between legitimate contract disputes and what she characterised as an overreach in this case. “The government could have simply canceled the contract as a less restrictive option,” she explained. “Labeling Anthropic a supply chain risk goes beyond the normal contract disputes seen in procurement negotiations with contractors.”

Huddleston also raised concerns about the international ripple effects of the dispute. If the US government successfully compels an American AI company to remove its safeguards, she argued, it creates a template that authoritarian governments elsewhere could point to when making similar demands of technology firms operating in their jurisdictions. “This raises significant concerns about potential for abuse both of the technology as well as interventions by governments to dictate the development or outputs of the technology,” she said.

On the relevance of the case to emerging markets, Huddleston said the lessons were directly applicable to countries designing their own AI regulatory frameworks. She argued that clear legal guidance on permissible government use of AI tools, including strong protections for civil liberties, should be established before rather than after disputes of this nature arise.

Anthropic has projected $14 billion in revenue for 2026, with the majority derived from businesses and government agencies using Claude for tasks unrelated to the Pentagon. The company has said the supply chain risk designation has a narrow scope, affecting only military contractors using Claude in work directly tied to the Department of Defense, but acknowledges the reputational and commercial damage from the dispute is already significant.