US expands federal push to speed AI infrastructure and set rules

In short: The US federal government is stepping up a long-running effort to grow AI by speeding up data center building, funding research, and pushing for clearer national rules.

What's going on

Over the past decade, the United States has used federal policies to support AI and related technology. That push has intensified since 2025, with a focus on making it easier to build the physical and legal foundations that AI needs.

In July 2025, the White House released an AI Action Plan and issued several executive orders. These orders aim to limit bias in government AI tools, speed up permits for new data centers, and encourage exports of US AI software and services. A data center is a large building filled with computers that power online services, similar to a power plant but for computing.

Federal agencies are also trying to increase capacity directly. The Department of Energy has asked for proposals to build AI-focused data centers on federal land, including a site in Paducah, Kentucky. On the private side, Amazon Web Services has said it will invest $50 billion to add 1.3 gigawatts of secure computing capacity for government AI work.

Funding is part of the plan too. Federal budget priorities for 2027 elevate AI and quantum technology, which is a newer kind of computing that can solve certain problems differently. Proposals include money for AI standards at NIST and funding for the National Science Foundation, including NAIRR pilots, which aim to give researchers shared access to computing resources (like a public library for computer power).

What to watch

A key question is how far Washington goes in overriding state AI laws to create one national approach, while still keeping exceptions like child safety. Another is whether faster data center approvals run into local concerns about land use, electricity, and water.

Source: Financial Times