Anthropic doesn’t trust the Pentagon, and neither should you

AI Summary

Anthropic, the maker of the Claude AI model, is engaged in an active legal battle with the U.S. Department of Defense after the Pentagon designated the company a 'supply chain risk' — a designation typically reserved for foreign technology suppliers suspected of embedding malicious tools. Anthropic responded by filing a lawsuit claiming the government violated its First and Fifth Amendment rights, arguing the designation amounts to an attempt to 'destroy the economic value created by one of the world's fastest-growing private companies.' According to The Verge's Decoder podcast featuring Techdirt founder and CEO Mike Masnick, the core dispute centers on two red lines Anthropic CEO Dario Amodei drew in contract negotiations: autonomous weapons and mass surveillance — with Anthropic specifically refusing to allow Claude to be used to analyze bulk third-party commercial data for surveillance purposes. The dispute is contextualized within a decades-long history of NSA legal reinterpretation, including the post-9/11 Patriot Act, Executive Order 12333 signed under President Reagan, and the FISA court system, which according to Masnick approved over 99% of surveillance applications and operated without adversarial oversight. In contrast to Anthropic's refusal, OpenAI's Sam Altman initially indicated willingness to accept 'all lawful uses,' a position Masnick suggests either reflected a misunderstanding of how the NSA has historically redefined statutory language — including key terms like 'target' — or a deliberate strategy to avoid public scrutiny, with Altman subsequently walking back that position. The Free Speech advocacy group FIRE has also entered the debate, publishing a blog post arguing that compelling Anthropic to build surveillance tools constitutes compelled speech under the First Amendment, an argument Masnick assessed as legally credible.