Why Building a Single AI Data Center Now Costs Tens of Billions of Dollars

Constructing a 1-gigawatt data center capable of supporting advanced AI workloads has reached an estimated $30 to $38 billion, a staggering expense that explains why technology companies face an unprecedented capital crunch as they race to deploy AI infrastructure. The costs reflect not only the sheer scale of modern facilities but also cascading supply chain disruptions, regulatory delays, and a fundamental shortage of electrical infrastructure in the United States.