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

The article discusses the legal battle between AI company Anthropic and the Pentagon, which designated Anthropic a supply chain risk, prompting a lawsuit alleging violations of the First and Fifth Amendments. The core issue revolves around the US government's history of surveillance, particularly how agencies like the NSA reinterpret plain English legal terms to expand surveillance capabilities, often without public scrutiny, as revealed by Edward Snowden. Guest Mike Masnick explains that this pattern involves government lawyers twisting interpretations of words like "target" to justify collecting communications involving US persons if a foreign person is mentioned, or collecting data that briefly leaves the US under Executive Order 12333. Anthropic's refusal to engage in mass surveillance, one of its stated red lines alongside autonomous weapons, contrasts with OpenAI's initial, seemingly more compliant stance. The situation is escalated because the Trump administration applied the "supply-chain risk" designation—usually reserved for foreign entities—to a US company for having an ethics policy, threatening to destroy its business. Furthermore, the discussion touches upon the Fourth Amendment and the 'third-party doctrine,' which allows the government to access vast amounts of personal data held by tech companies (like cloud providers) with less stringent warrant requirements, effectively undermining constitutional protections.

(Source：The Verge)