Court Rules Trump Administration’s Blacklisting of Anthropic Was Unconstitutional

A federal court has ruled that the Trump administration’s blacklisting of AI company Anthropic was unlawful and unconstitutional, finding that the government retaliated against the company for protected speech related to AI safety and military use.

U.S. District Judge Rita Lin said the Pentagon’s designation of Anthropic as a “national security supply-chain risk” violated the First Amendment. The decision followed Anthropic’s refusal to remove restrictions preventing its Claude AI from being used for fully autonomous weapons and mass domestic surveillance. The Verge reported on the legal fight and the government’s effort to cut Anthropic out of defense contracting.

The ruling matters because it places limits on how national-security powers can be used against a domestic technology company. The court also found due-process concerns, saying Anthropic was not given a meaningful opportunity to challenge the designation before facing serious economic consequences.

For everyday users, the case highlights a larger question: who decides how powerful AI systems may be used: and what protections should companies have when they object to government policy? The answer could influence future rules for workplace AI, government contracts, privacy, cybersecurity, and automated decision-making.

The news arrives as OpenAI and Anthropic join more than 100 organizations in an international call for stronger defenses against AI-powered cyberattacks. The signatories are urging governments, technology companies, and critical infrastructure operators to share threat intelligence, strengthen access controls, fund defensive tools, and prepare for attacks that can move faster with the help of autonomous AI agents.

As France 24 has reported, AI agents can take actions online, access sensitive accounts, and potentially follow hidden malicious instructions. Residents and small organizations can respond by using multifactor authentication, updating software, limiting account permissions, and treating AI-generated code or messages cautiously. Brownstone’s earlier coverage of AI agents and cybersecurity offers additional context.

Should AI companies be allowed to set firm safety boundaries: even when governments disagree? Share your perspective with the Brownstone Worldwide community.