Fortune just reported that the US alone has 1,200 AI bills and no good test for any of them, because “in the past nine months, the United States has produced more AI legislation than in the prior decade, [that faces] three different [and partly contrasting] theories of what AI policy is supposed to do”.

The lack of clarity coming from that proliferation of rules, said IBM Chairman and CEO Arvind Krishna, will be the “central question facing AI policy”, because:

1. “The balance between too many regulations, it’s terrible; too few, we may not love the outcome, so we got to find the Goldilocks middle.”

2. Whatever it is, that Goldilocks middle must work both inside the US and abroad, because if the international regulatory landscape turns into a bloated bureaucracy, “that would not be so good for us to win the AI race.”

Is that really the problem, or something else? Never mind

Of course, all those concerns come from one explicit pretension and constitute one implicit, but desperate request: the pretension is that AI for the whole world must be 100% made, deployed and ruled the American Way. Even if, and that’s the request between the lines, the American AI industry is on the verge of the greatest bankruptcy ever, please don’t make it even faster with too many laws!

That’s no news, of course. This AI race will crash big, the sooner the better, but never mind that open secret now. That Fortune article proves me right about something else, and it does with these statements (emphasis mine):

• “By one count, state legislatures introduced over 1,200 AI-related bills in 2025 and enacted just under 150, with the pace accelerating since.”

• “Beneath the volume lies a more fundamental problem. Policymakers at every level are working without a shared test to determine whether their legislative efforts constitute good policy.“

• “State bills creating new accountability rights for automated hiring often duplicate protections already enforceable under [other American laws].”

• “The legislative volume is high, but a shared test for distinguishing good policy from bad has been absent from the debate.”

The piece then goes on to propose a testing framework that “will not, on its own, resolve any specific dispute. Its purpose is to ensure that the questions before state legislators, members of Congress, and federal agencies are the right questions, asked in the right order, before another five hundred bills are introduced and a patchwork is hardened in place that no one designed and few defend.”

Boy, am I happy. Thanks for proving me right!

This mess, which of course isn’t limited to the US, is exactly what I was talking about one week ago, when I wrote:

Imagine if there was [some AI] that, every time a new law proposal is published, everybody could use to easily discover that that proposal violates, contradicts or duplicates in any way other, already existing laws.

Imagine if there was [some AI] that, after reading all the laws of a country, produced a completely equivalent but much shorter, much more readable set of laws. Something that would make it much quicker and easier for everybody to see what’s good and bad in every law, and for whom.

Dear AI companies: if your AI is as good as you claim...

Make your AIs do on all those AI bills just the job I suggested. Make them predict and describe all the cases in which all those hundreds of bills will contradict or duplicate each other, or fail to address some issues.

Make your AIs be the paralegal that does all the boring job that humans, of course, will need to have done before they can start to “determine whether their legislative efforts constitute good policy.” If your AIs can find bugs in AI bills as I suggested last week, it will be great advertising for your products. If it fails, no problem. At least it won’t make anyone aim missiles at schools. What do you have to lose?

So, make AI do all those comparisons. Publish the full result online, in the Public Domain, so everyone can evaluate its quality, with a bounty program for finding errors. Tell all the 120K+ law students in the US that you’ll pay all the present and future educational debt of every student that finds at least one bug.

And of course... don’t limit that work to the USA. Make your AIs repeat the whole program in every country, bounties and all, because every country has the right to get its own AI laws that conform to its own ideals, needs and interests.

Seriously: you have flooded us for years with claim after claim of how superintelligent and super capable and omniscient your models are, so much that they could even kill us all… and I must be the one who has to suggest to use those same models to figure out in the laws that would hinder them?

As usual… Thanks in advance for your subscription (possibly but not necessarily paid) or, even better, donations for any amount, or consulting, writing or speaking work.

Subscribe now