Pentagon Flags China-Linked Staff as Commercial AI Is Deeply Embedded in U.S. Military Systems

Lawsuit spotlights national‑security anxieties

The latest courtroom fight between Anthropic and the U.S. Department of Defense has put a sharp spotlight on a sensitive question: how much of Silicon Valley’s talent — and whose loyalties — feed the military’s rapidly expanding use of commercial AI? Anthropic sued the Department after it faced a “supply‑chain risk” ban and sought an emergency injunction; the Pentagon has filed an opposition that the Department of Justice will argue in court next week. Emil Michael, the Pentagon’s deputy for engineering and technology, filed a sworn declaration saying Anthropic employed a large number of foreign nationals and singled out employees from China as a potential security risk.

Silicon Valley’s deep military ties

The affidavit underscores a broader reality: U.S. defense projects now routinely integrate commercial large language models, cloud services and GPUs from the private sector. Firms named in public accounts include Palantir, Microsoft, Amazon Web Services, Google/DeepMind, OpenAI, Anthropic and xAI; Nvidia’s accelerators are widely described as the physical backbone of military AI. It has been reported that Anthropic’s Claude has been used in sensitive operations and that OpenAI and xAI have since won Pentagon contracts; similarly, U.S. venture capital — from a16z to Founders Fund and In‑Q‑Tel — has poured billions into defense‑adjacent startups. At the same time, roughly 40% of Silicon Valley’s AI talent is estimated to be Chinese‑born, raising fraught questions about access, vetting and trust. Who polices that line — industry, government, or both?

Geopolitics, personnel departures and what’s at stake

The dispute is unfolding against an intensifying geopolitical backdrop: trade restrictions, export controls and “supply‑chain risk” designations are now routine levers of U.S. policy toward China, and tech companies find themselves squeezed between commercial opportunity and national‑security scrutiny. It has been reported that several high‑profile Chinese‑born engineers have left or been pushed out of AI startups such as xAI (founders and core staff including Greg Yang (杨格), Tony Wu (吴宇怀) and Jimmy Ba (吉米·巴) are cited in media accounts), a development that has stoked controversy and raised questions about workplace dynamics and compliance. More details will emerge in court before Judge Rita F. Lin (林萍), the first Chinese‑American woman to sit on the U.S. District Court for the Northern District of California — and the outcome could shape hiring, contracting and the strategic calculus linking Silicon Valley to the U.S. warfighting enterprise.