Post by Chris Dover

AI-First Founder-Engineer | Agentic Systems, Defense Tech, Trading Systems | 500 Startups Alum

The data center guard shack had bullet holes in it. Not overseas. Not in Latin America. Not in any of the high-threat environments where I have spent most of my career. This was a US data center. One of the big hyperscalers. Billions in infrastructure. The most advanced water-cooled GPU farm I have ever seen sitting behind the fence. Outside the fence, a local fringe group had been ambushing construction workers. I counted between five and ten drones overhead at any given time. Not all of them belonged to the project. I have managed security operations in high threat environments all over the world. I know what a contested environment looks like. That data center was one. The difference is that overseas, organizations plan for this. In the US, the assumption is that domestic means safe. It does not. The AI industry is spending billions on infrastructure and almost nothing on the physical safety of the people who build and operate it. No threat assessments. No personnel security programs. No counter-surveillance. A guard service contract and a camera system protecting assets that nation-states are actively trying to compromise. If you sit on a board, serve as General Counsel, or run security for an organization building AI infrastructure in the United States, and nobody with operational security experience has walked your sites and told you what they actually found, that is the conversation you need to be having this week. I wrote about it. Link in the comments.

Post content