US essay urges shared state–federal compact to govern AI growth

Beier’s essay is a window into how the US legal establishment is starting to normalise AI as critical infrastructure, akin to railroads or electricity, that demands shared governance between Washington and the states. Rather than arguing for a single, sweeping federal AI law, he sketches a division of labour: the federal government handles national security and cross‑border risks, while states retain control over data‑center impacts, utilities and traditional tort law around AI harms. ([theregreview.org](https://www.theregreview.org/2026/01/12/beier-a-new-state-and-federal-compact-for-artificial-intelligence/))

For the race to AGI, the significance is subtle but real. If this view gains traction, frontier labs may face a patchwork of state‑level rules on energy use, water, siting and liability even as they lobby for lighter federal oversight. That could slow the rollout of massive data‑center projects in some jurisdictions while accelerating them in states that position themselves as AI‑friendly. It also hints that any future federal AI safety regime will likely layer on top of, not sweep away, emerging state laws targeting large‑scale systems.

In competitive terms, companies with the resources to manage multi‑jurisdictional compliance—big US tech platforms and well‑financed labs—will have an edge over smaller players. But the flip side is that states retain leverage to insist on transparency and local benefits when hosting the infrastructure that will power AGI‑class systems.