AI regulation: Europe vs America

The EU AI Act is now enforceable. The US still has no federal AI legislation. Let's compare the two approaches and what they mean for the industry.

The EU approach is precautionary: regulate first, ask questions later. The US approach is permissive: innovate first, regulate when problems emerge. Both have real costs.

The equity dimension gets lost in this debate. AI systems are already making decisions about who gets loans, who gets hired, who gets paroled. Unregulated AI disproportionately harms marginalized communities.

The trade dimension is significant too. If EU and US regulatory frameworks diverge, companies face compliance costs that effectively fragment the market. This is the new non-tariff barrier.

The state-level patchwork in the US is the worst of both worlds. Colorado has an AI act, California has draft legislation, and they're not compatible. Companies are begging for federal preemption just to have one set of rules.

From a constitutional perspective, the question is whether AI regulation is a commerce clause issue or a state police power issue. The answer determines whether Congress or statehouses set the rules. This will end up at the Supreme Court.