More than 400 pages released under a FOIA lawsuit show the Pentagon bought strategy advice and adversary briefings with the models, and OpenAI staff inside its commands.
The Department of War put OpenAI's ChatGPT Mil and Starshield AI's Grok on GenAI.mil the same day, and the limits each vendor accepts now differ more than the models do.
A Reuters review of more than 80 Chinese papers and patents found PLA-linked researchers training domestic defense AI on outputs from OpenAI and Anthropic models, a transfer route Washington's chip export controls do not cover.
Sen. Elizabeth Warren gave the Pentagon and seven AI vendors until July 20 to open their classified-network AI agreements, weeks after Senate defense authorizers wrote human-judgment requirements and a nuclear-launch ban for military AI into their draft policy bill.
Five Senate bills in two weeks would write human control over autonomous weapons into statute, the legislative answer to Trump's order to speed military AI adoption.
A new national security memo tells the Pentagon to onboard frontier commercial AI at speed, strips vendors of the power to shut it down mid-mission, and resets the rules for autonomous weapons within 90 days.