The most consequential AI regulation in America this year did not come from Washington. It came from Springfield. When Illinois Governor J.B. Pritzker signed a landmark bill aimed at mitigating AI risks, he did something more significant than adding one state to a list: he completed a bloc. Illinois, California and New York now share a broadly similar regulatory posture, and lawmakers behind the Illinois law estimate the three together account for roughly 40% of the US AI market. That figure is the whole story. In the absence of federal rules, a handful of large states can quietly become the standard everyone else builds to.
This is not how most people imagine AI gets governed. The dominant picture is top-down: a sweeping federal act, or an international treaty, dictating terms from above. The reality is messier and more revealing. AI is being regulated by whoever moves first at scale — and that has lessons far beyond Illinois, including for India.
What was signed
According to WTTW News, Pritzker signed a landmark AI regulation bill explicitly designed to mitigate the risks associated with automated and generative systems. Its architecture borrows heavily from earlier laws in California and New York, a deliberate choice that matters as much as the substance. By modelling its statute on existing frameworks rather than inventing a fresh one, Illinois reduced friction for companies already adapting to those two states — and reinforced a shared template that others can copy.
The bill is a response to a vacuum. For all the noise in Congress about AI safety, meaningful federal legislation has stalled. That stall is not neutral; it is a decision by default that leaves the field open. States, which face the harms of AI directly — in hiring tools, insurance decisions, deepfakes, consumer scams and public-sector automation — have far less appetite to wait. Illinois, like California and New York before it, chose to act on the problems in front of it rather than defer to a federal framework that may be years away, if it arrives at all.
Why states set the standard
Here is the mechanism that turns a state law into a national one. Complying with the strictest rule in a large, unavoidable market is usually cheaper and simpler than maintaining different versions of a product for every jurisdiction. As policy analysts noted around the Illinois signing, that economic logic means first-mover state laws tend to become national norms by default — shaping how AI is governed well beyond the borders that enacted them.
Consider the incentives. A company selling an AI hiring tool or a content-generation platform cannot easily ship one compliant version to California and a looser one to a neighbouring state — not when California, New York and now Illinois collectively represent something close to 40% of the US market by lawmakers’ own estimate. The cost of building, testing and maintaining multiple compliance regimes typically exceeds the cost of simply meeting the highest bar everywhere. So firms tend to standardise upward.
The result is a de facto national standard that no national body ever passed. This is the same dynamic that gave the world California’s vehicle emissions rules and, arguably, the European Union’s data-protection reach. Regulation flows toward the largest market with the strictest rule. Once three of the biggest AI markets in the country converge on a similar template, the remaining states and the companies serving them are effectively negotiating with a rulebook already written.
The trade-offs
None of this is unambiguously good. A patchwork built from whichever states legislate first is not the same as a coherent national policy, and the differences carry real costs.
- Fragmentation risk. Three aligned states is convergence; a dozen mismatched ones is chaos. If other large states pass laws that conflict rather than harmonise — different definitions of high-risk AI, different disclosure duties, different enforcement triggers — companies face genuine confusion rather than a single high bar. The tidy “comply with the strictest” logic only works while the strict rules point in the same direction.
- Compliance costs and who bears them. Large incumbents can absorb legal and engineering overhead that startups cannot. A well-intentioned law can, in practice, entrench the biggest players by making it expensive to enter the market. The upward standardisation that protects consumers can also raise the drawbridge behind established firms.
- Democratic mismatch. When Illinois, California and New York effectively set rules for the country, voters in the other 47 states inherit standards they never elected anyone to write. That may produce good outcomes, but it is governance by market gravity, not by mandate — and it is worth being honest about the difference.
The counterweight is that the alternative — waiting for a perfect, comprehensive federal framework — has so far delivered nothing. A functional patchwork that leans toward protection may beat an elegant policy that does not exist. The honest position is that state-led standard-setting is a second-best solution doing first-best work, and its legitimacy depends on states coordinating rather than colliding.
The India read
For readers in India, the American story is not a curiosity — it is a preview of how global standards actually form. Sub-national and first-mover rules ripple outward. Indian firms that sell to US customers, or that build on models trained under these regimes, will inherit constraints shaped in Sacramento, Albany and Springfield long before any Indian statute touches them. The practical takeaway for founders and operators is to watch where the strictest large market lands, because that is often where the real product requirements come from.
India’s own approach to AI governance has been deliberately lighter-touch, favouring guidance, advisories and sectoral oversight over a single omnibus AI law. The bet is that premature, heavy regulation could stifle a fast-growing domestic ecosystem. There is merit in that caution — but the Illinois lesson complicates it. Standards are being set with or without India’s participation. Sitting out the rule-writing does not mean escaping the rules; it means importing someone else’s.
The more interesting opportunity is for India to do at national scale what the US is doing by accident across states: design rules that protect without stifling. A large, unavoidable market that writes clear, proportionate AI standards — strong on transparency, discrimination and high-risk uses, restrained on the ordinary tooling founders depend on — could become a template exporter in its own right, not just a rule-taker. India has the market weight to make that credible. The question is whether it uses it.
The deeper lesson of the Illinois law is about power, not policy detail. AI regulation is not descending from some central authority with a grand plan. It is being assembled by whoever is large enough and fast enough to move first. That reality rewards jurisdictions willing to lead and quietly binds those that hesitate. For everyone building in this space — in Chicago, Bengaluru or anywhere in between — the rulebook is being written right now, by the first movers at scale. It pays to read it early.
