The Pentagon has 1.7 million AI users but weapons programs running 12 years late — today's briefing unpacks that contradiction, plus Mistral's €11.7B Series C and Europe's sovereignty play. Fast, sharp, and essential for anyone tracking AI's collision with defence and geopolitics.
Audio is available on Spreaker — see link below.
The Pentagon now has one point seven million people using its internal AI platform. That's the headline.
The GAO's latest report found that one hundred and four major defense programs average more than twelve years behind their original delivery schedule. Twelve years.
There's one area where the Pentagon is trying to use AI to fix the bureaucracy itself. The Department of Defense is piloting AI agents to automate the authority-to-operate compliance process, the two-year approval cycle required before any software system can be deployed on defense networks.
Meanwhile, the workforce picture is deteriorating. The Defense Department lost more than twenty-four thousand technical employees in fiscal year twenty twenty-five.
On the other side of the Atlantic, France's Mistral AI is making a different kind of bet. The company closed a Series C round at a valuation of eleven point seven billion euros, backed by an ASML-led consortium, and it's committed four billion euros to building data center infrastructure in France and Sweden.
The two watchpoints that matter most right now are these. First, whether the Pentagon's AI agent pilots for ATO compliance produce real cycle-time reductions or just faster paperwork.
Chapter summary auto-generated from the verified script. Listen to the full episode for the complete content.