Cargo theft targeting AI data center hardware has escalated dramatically. Thieves hijacked two armored shipments traveling from Silicon Valley to Southern California in recent months, using tactical maneuvers—PIT (precision immobilization technique) maneuvers and rear-end collisions—to disable security escorts, then fled with multi-million-dollar server loads. Neither truck has been recovered. Investigators believe the truck drivers were complicit, with both incidents involving motor carrier numbers that recently changed hands, a known fraud technique.
The broader theft landscape shows Q2 2026 supply-chain theft losses hitting $304.6 million across the U.S. and Canada—more than double the $135.7 million in Q2 2025—despite incident volume dropping 26% year-over-year. Average theft value surged to $564,009. Enterprise computing and networking equipment ranks among the most targeted categories. Verisk CargoNet documented 677 incidents in Q2, down from 910 in Q2 2025, but the per-incident value skyrocketed.
The economics of AI hardware theft have inverted due to U.S. export controls: restricted Nvidia servers now sell in China at steep premiums over U.S. retail. A Nvidia DGX B300 server—an unrestricted part, but highly sought by labs in China—commands $1.1 million on the black market against a U.S. retail price of $550,000. Even five-year-old A100 servers have tripled in price. A hijacked truckload of restricted AI hardware is now worth more on the black market than off it.
For supply chain officers and datacenter operators shipping Nvidia or other restricted hardware, the threat profile has shifted. Organized cargo theft rings are explicitly targeting AI infrastructure; location pings suggest many shipments end up overseas within weeks. Freight security costs, routing changes, and insurance adjustments are now non-optional.