Governments and Tech Giants Treat AI as Critical National Infrastructure
The latest State of AI report highlights how global governments and tech giants are building massive compute capacity, dismissing ongoing bubble narratives. Major investments from the US Department of Energy and NVIDIA underscore the shift toward treating AI as essential sovereign infrastructure.
The November 2025 State of AI newsletter pushes back against recurring bubble narratives by pointing to real enterprise adoption and massive revenues at hyperscalers and AI labs. Even if there is an overbuild in infrastructure, the newsletter argues that this primarily drives down costs and creates durable physical assets like semiconductor fabs and power plants rather than signaling an impending market collapse.
Governments worldwide increasingly treat AI compute capacity as a matter of national security and sovereignty. In the United States, the Department of Energy partners with AMD to build new "AI Factory" supercomputers at Oak Ridge National Laboratory. These new systems, named Lux and Discovery, carry a roughly $1 billion combined public and private budget and are scheduled to come online in 2026 and 2028.
Private tech giants mirror this government push with their own massive infrastructure investments. NVIDIA makes a $1 billion investment in Nokia to strengthen US communications infrastructure, reflecting the same underlying logic of scale that drives federal AI initiatives. Additional developments include Air Street Press highlighting Poolside's entry into AI power infrastructure and a new £2 billion partnership with NVIDIA to accelerate the UK's AI ecosystem.