Forcing AI Makers To Legally Register Their AI With The U.S. Government Stirs Intense Reactions

AI And The Law

As a quick background, I’ve been extensively covering and analyzing a myriad of facets regarding the intersection of AI and the law for many years. You can find my writings not only in my Forbes column but also as posted in Bloomberg Law, ABA Law Journal, The National Jurist, The Global Legal Post, Lawyer Monthly, The Legal Technologist, MIT Computational Law Journal, and so on.

There are two major perspectives on the mixture of AI and law:

• (1) Law & AI. The application of laws to the governance and regulation of AI.
• (2) AI & Law. The application of AI to perform legal reasoning.

Thus, you can apply the law to AI, and conversely, you can apply AI to the law. For my big picture overview of both of these exciting and rapidly evolving realms, see my discussion at the link here and the link here.

When it comes to applying the law to AI, the aim is to establish suitable regulations and provide appropriate governance on how AI should be devised and implemented. There are longstanding concerns that AI makers aren’t giving due attention to the ethical ramifications of their wares. Ethical issues are construed as “soft laws” and aren’t as formidable as legally enacted laws, known as “hard laws”. To level the playing field and keep AI makers on the up-and-up, some believe that we need more AI laws.

On the other side of the coin is the application of AI to the law. This consists of using AI to aid legal activities. Lawyers tap into the latest AI to devise legal strategies, brainstorm to find creative legal arguments, draft court filings, and prepare for cases by having the AI pretend to be an able adversary. For my extensive coverage on AI for legal reasoning (AILR), see the link here.

The Current Situation Legally

In terms of the AI laws in the United States, they have not yet stood the test of time, meaning that we won’t really know how well they stand up until there are court cases that test these new laws. It is too early to know whether the laws will survive legal battles waged by AI makers and other contenders. Just because AI laws are enacted does not mean they are proper. All sorts of improper provisions and constitutionally contentious stipulations are undoubtedly buried within these shiny new AI laws.

Congress has repeatedly waded into establishing an overarching federal law that would encompass AI. So far, no dice. The efforts have ultimately faded from view. Thus, at this time, there isn’t an overarching federal law devoted to these controversial AI matters. The big question will be to what degree a sweeping federal law would impact the numerous state-level AI laws. The odds are that many state-level laws would run afoul of a federal mandate, and a tsunami of legal cases would arise as a tussle between federal and state law is undertaken. It surely will be a legal mess.

The crux is that there is intense and pervasive interest in using the law to govern AI. It is an abundantly burgeoning realm. AI companies would be wise to keep a close eye on what is happening in the hallways and byways of regulators and legislative bodies. I have repeatedly noted that a profitable specialty for budding lawyers is to consider concentrating on the exciting and dynamic field of AI and the law; see my predictions and suggestions at the link here.

Difficulties Aplenty

You can likely envision the challenges of the legal landscape governing AI.

Each state does its own thing. The AI laws in some states are poorly specified and legally ambiguous. States are also amending their AI laws that they previously thought were perfect. Other states that haven’t been enacting AI laws are opting to jump into the waters with both feet. They might borrow wording from other states, change it up, and put it into their legal books. Estimates suggest that there are well over 1,000 AI-related bills and laws that are in some form of consideration at the state level, ranging from pending status to actual enactment.

I’ve been extensively analyzing and explaining the disparate and at times conflicting state-level AI laws; see the link here. There are plenty of downsides to this situation. Plus, the matter is worsening. Public interest in AI laws is heightening. State-level lawmakers are becoming more familiar with AI and are joining the bandwagon on laws about AI. All told, a grand convergence is taking place toward a veritable tsunami of new AI laws across all 50 states.

Debating The Registration Of AI

Now that you are familiar with the overall legal landscape associated with AI in the United States, we can turn to the momentous issue that is being batted around by policymakers and lawmakers:

• Should AI makers be required to officially register their AI with the U.S. government?

The idea is somewhat straightforward. Just like a car needs to be registered, so too should AI be registered. An AI maker would file some form of paperwork with the U.S. government that details the nature of their AI. This then would formally signify that they have dutifully registered their AI. It would be on record, and we would know which AIs have been registered and which AIs have not yet been registered.

Seems easy. No fuss, no difficulty. Just have the government set up an online database to keep track of AI registrations. AI makers would create an official account and then start registering their AI. If an AI maker discontinued an AI that they had previously registered, they would access the database and mark that their AI is now defunct. When an AI maker comes out with a new version or model of their AI, they would need to update the database.

The Litany Of Thorny Problems

Of course, there is no such thing as a free lunch. I mention this because the seemingly simple aspect of having AI makers register their AI with the U.S. government has a litany of thorny problems. There is plenty of debate and quite disparate viewpoints on who, what, where, why, when, and how of such a scheme.

Let’s handily consider these ten key factors:

• (1) Scope. What AI is construed as within the scope of the registration requirement?
• (2) Responsibility. Who is to be held responsible for registering AI that falls within the scope of the registration requirement?
• (3) Validation. In what way is the AI registration to be sufficiently validated?
• (4) Timing. When does the AI that falls within the scope have to be registered?
• (5) Jurisdiction. Where geographically or jurisdictionally does the AI have to be made or be utilized to fall within this registration scheme?
• (6) Disclosure. How is the information that is provided during registration to be disclosed?
• (7) Compliance. If AI is not registered, what happens, and likewise, if AI is registered, what does this allow?
• (8) Updating. When is AI registration to be updated, and what penalties are imposed for failing to do so?
• (9) Federal. Which federal agency or outside entity would be responsible for the operational aspects of the registration scheme?
• (10) Exemptions. Are there AIs that fit within the rubric but are allowed to be exempt from registration?

I rank these factors as the ten essentials. Please know that many more factors can be considered. Also, since each of those factors is quite complex, I’ll just briskly cover them overall and will be doing a series of postings that go into great depth for each factor. Stay tuned.

Scope And Responsibility

The first place to start would be by asking the explicit question about what constitutes a semblance of AI that needs to be registered. Would everything that could conceivably be construed as AI be within the scope of artifacts that must be registered? If so, this would be not only voluminous but overly egregious. There must be some kind of definitional scope associated with the nature or type of AI that would fall into this rubric.

A frequent assertion is that the AI would have to be so-called frontier AI, also often noted as consisting of AI foundational models (see my detailed explanation at the link here). This is a nebulous term. It generally means that an AI is at the frontier of the latest advances in AI and often is the largest AI around. That doesn’t, though, provide sufficient clarity, and we would once again have confusion over what lands in this sphere. Sometimes the threshold is based on the number of parameters, dataset size, FLOPs, compute consumed during training, etc. A criticism is that size alone should not be the determining ingredient since even a small AI can pose great risk and danger.

The party responsible is also a muddled aspect. You might say it is obvious: the AI maker that made the AI is the responsible party. Suppose an AI developer crafts an AI that doesn’t fit with the registration scope, perhaps it is below the required threshold. Fine, they don’t have to register the AI. A second AI developer comes along and expands the AI, which then falls within the scope of registration. Is the second AI developer the responsible party for registration, while the first AI developer is still off the hook?

Validation And Timing

An AI maker opts to register their AI. Who checks the registration to ensure that it is truthful and complete? One perspective is that no one checks it, but if later the registration is shown to be false or incomplete, the AI maker is subject to criminal and civil penalties. The problem there is that for some period of time, the registered information is sitting there and assumed to be valid. The public would be misled.

Maybe validation should be undertaken right away. This would require the consumption of resources. Would the government do this? Would some third party be hired and paid to do this? Would the registering party pay for this, or would it be covered by the government? The cost aspects of the entire registration scheme would need to be carefully worked out. If the cost is exorbitant, it might put smaller AI makers into a bind and essentially serve as a form of regulatory capture, favoring large AI makers that can afford the costs of registration.

At what point in time does the AI maker have to register the AI? You might argue that only once the AI is released into public use does the registration have to occur. Up until then, the AI was merely in-house. The downside there is that even while AI is in-house, the testing process can lead to the AI breaking out and causing issues (see my coverage of such an example, the Mythos AI, at the link here). Perhaps the registration should be undertaken as soon as the AI is designed, or maybe once it is in a testing mode. Lots of possibilities.

Jurisdiction, Disclosure, And Compliance

One belief is that any AI maker that is based in the United States would potentially be a candidate for registering their AI. In a sense, this gives foreign AI makers a leg up, namely they do not have to deal with the registration scheme. Yet their AI might be fully accessible by those inside the United States. Perhaps a U.S.-based AI maker would relocate to a foreign country to escape the registration process. The jurisdictional issues are challenging.

Next, mull over what type of information an AI maker should be required to register concerning their AI. The disclosure could compromise their proprietary capabilities. Is that fair to the AI maker? Maybe the disclosure is kept secret, such that only the government knows what the registration contains. That doesn’t seem to be useful to the public at large. Some argue that part of the disclosure should be private and other portions would need to be made publicly available.

A registration scheme must have teeth to garner compliance. AI makers are unlikely to comply if they don’t see any downside to not registering. How would the government know when an AI maker avoids registering? What consequences are there? An AI maker might opt to drag the matter out in court and avert being registered for years on end.

Updating, Federal, And Exemptions

An AI maker registers their AI dutifully with the governmental registration system. A few months later, the AI maker updates their AI. Should the AI maker be obligated to change or update the registration status? If so, what is the deciding line? An AI maker might claim that the changes were mild and do not warrant having to update their registered aspects.

Which governmental agency is going to establish, monitor, and report on the AI registrations? It could be an existing entity. Maybe a new entity should be created for this purpose. Some believe that the government should not have a hand in it at all. They argue that the only role of the government would be to pass laws about the matter. The registering entity would be a non-profit or some other form of private organization.

Should exemptions to AI registration be allowed? For example, a university creates an AI that fits within the rubric of the AI registration scheme. Some might contend that a university shouldn’t have to register the AI; they deserve an exemption. Others decry the idea of exemptions and flatly declare that no matter what the circumstance, if the AI fits the rubric, it must be registered.

Choosing The Choices To Be Made

I’m sure that you can now plainly see that there are lots of options and paths involved in pursuing AI registration. Different policymakers, lawmakers, researchers, AI developers, and the like are bound to have disparate views. When you see some strawman indication of an AI registration scheme, make sure to weigh it against the ten factors and identify what postures or positions the scheme has chosen.

If you are interested in the layout of an AI registration approach, you might want to look at a recently published paper entitled “Legal Infrastructure For Transformative AI Governance” by Gillian K. Hadfield, PNAS, July 20, 2026, which made these salient points (excerpts):

• “Automobiles must be registered with a department of motor vehicles and issued a unique vehicle identification number (VIN).”
• “Against this history and the essential role of registration to underwrite the effective functioning of regulatory regimes in the modern economy, it is striking that there are no registration regimes in place specifically with respect to AI in the United States.”
• “The technologies of the most advanced models are still being built largely inside private companies, protected from public view by trade secret law, confidentiality agreements, and employee obligations to maintain company secrets.”
• “Registration is an important legal regime that allows us to build appropriate and effective substantive AI regulation.”
• “Disclosure should be reasonably comprehensive and subject to penalties for lack of completeness or fraud, but ideally this should not be a major burden on developers, large or small.”

An element in the above-cited paper is that the registration not only applies to the AI makers but also to the buyers of AI. Think of this in the context of buying a car. You are expected to check the VIN of the car and ensure that it is a valid VIN and also not a stolen car. The same kind of logic might apply to those who buy or make use of AI services. Is that onerous, or does it provide a prudent check-and-balance on AI registrations?

The World Ahead

A worry some have is that people will mistakenly assume that just because an AI is registered, that it signifies some kind of certification of safety. Aha, people will be thinking, I will use that AI over someone else’s AI that isn’t registered. But the AI registration schemes typically have nothing to do with testing or certification. Maybe this should be the case, though that would exponentially create a mountain out of a molehill (i.e., a firestorm debate would ensue, and we might not ever land on even the most basic of registration processes per se).

A final thought for now. The famous Roman statesman Marcus Cicero made this insightful remark: “The safety of the people shall be the highest law.” Do you think that the safety of us all would be enhanced or would it be reduced if AI makers had to officially register their AI? It’s the kind of vexing question that will take great collective wisdom to resolve.