Anthropic, the AI safety company behind Claude, has filed a lawsuit against the US government challenging its designation as a supply chain risk. The label, applied under the Trump administration, effectively restricts the company's ability to work with federal agencies and defence contractors.

The designation followed a period of tension between Anthropic and the White House. Unlike OpenAI and Microsoft — who both made substantial commitments to the administration's AI agenda — Anthropic declined to offer assurances that its models would be made available for autonomous weapons targeting and mass surveillance applications. That refusal, according to the company, is the real reason for the label.

What the Designation Means

Being classified as a supply chain risk under US national security frameworks is not a minor inconvenience. It signals to government contractors, federal agencies, and partner organisations that working with Anthropic carries national security implications. In practice, it can freeze out a company from large swaths of government and defence-adjacent work.

For a company that has positioned AI safety and responsible deployment at the centre of its identity, the irony is considerable: being punished for refusing to compromise its safety standards.

The Political Context

Senator Kirsten Gillibrand has been among those to speak out, calling it deeply troubling that a government would blacklist a company specifically for refusing to enable autonomous lethal systems. The designation came after Anthropic's CEO Dario Amodei notably did not donate to Trump's inauguration fund — in contrast to Sam Altman of OpenAI, who contributed $1 million.

Meanwhile, Anthropic's user base continues to grow regardless. The company reports adding approximately one million new Claude users per day. Mayor Sadiq Khan has separately written to invite the company to expand operations in London — a signal that whatever Washington thinks of Anthropic, other governments are happy to welcome it.

The lawsuit sets up a significant legal test: can a government designate a private technology company a security risk primarily because it refused to compromise its ethical standards?

Source: BBC News