Reverse Federalism

Reverse federalism is the idea that when there's no nationwide rule yet, building out a solid regional rule first can end up becoming the blueprint for the eventual national rule.

Usually it works the other way around. The central government sets the big framework first, and each region only adjusts the details within that framework. But when the central government takes too long to set the rules, regions start making their own different rules. It's a bit like a school where each classroom makes its own rules because there's no school-wide rule yet. Once that happens, merging all those different classroom rules into one later becomes harder.

So the alternative that emerged is: if state-level rules are going to appear anyway, align and refine them from the start. If one state's rule is made well and precisely, other states—or a future federal rule—can reference it. For companies being regulated, it's also easier to work off one well-crafted standard than to comply with 50 completely different state rules.

As OpenAI pushed for strengthening California's AI safety law SB 53, it said it supports this kind of 'reverse federalism' approach given the absence of clear federal AI legislation. What's easy to misread here is that this doesn't mean OpenAI welcomes regulation itself. Rather, it's more of a strategic calculation—to avoid the burden of complying with a patchwork of different state rules, OpenAI wants California's rule refined well enough to serve as the starting point for other states' and the federal standard.