In today’s column, I examine a frequent mistake that lawmakers are making when it comes to regulating AI. Lawmakers tend to miss the boat by failing to call out that there are AI apps that are purpose-built versus AI apps that are of a general-purpose nature.

Here’s why this is important. Consider the realm of AI and mental health. People use general-purpose AI such as ChatGPT, GPT-5, Claude, Grok, Gemini, CoPilot, and other LLMs to get mental health advice. That generic type of AI is not tailored for psychological guidance. Meanwhile, there are specialized AI apps that are purpose-built to provide mental health support. This is their primary and often sole purpose.

When devising new laws for regulating AI, lawmakers need to clearly indicate whether the AI being governed is purpose-built versus general-purpose AI. This provides vital clarity. For example, does an AI law on mental health apply to purpose-built AI, general-purpose AI, or both? By not giving explicit consideration to the distinctions, the chances are that the budding law will have numerous troubles downstream. The law will be confusing, confounding, and create potential legal loopholes that allow AI makers to skate free.

Let’s talk about it.

This analysis of AI breakthroughs is part of my ongoing Forbes column coverage on the latest in AI, including identifying and explaining various impactful AI complexities (see the link here).

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.

Purpose-Built AI Versus General-Purpose AI

On the matter of purpose-built AI versus general-purpose AI, I have identified five major flaws in the burgeoning realm of AI laws:

(1) Purpose-built AI Myopia. A law focuses only on purpose-built AI and overlooks general-purpose AI.
(2) General-purpose AI Myopia. A law focuses only on general-purpose AI and overlooks purpose-built AI.
(3) Misses The Mark. A law uses legal stipulations that apply to purpose-built AI, but the law is focused on general-purpose AI, or the legal stipulations apply to general-purpose AI, but the law is focused on purpose-built AI.
(4) Inapt Lumping. A law lumps together inaptly both purpose-built AI and general-purpose AI.
(5) Failure To Specify. A law is nebulous and doesn’t stipulate whether purpose-built AI and/or general-purpose AI is the intended aim of the law.

Let’s briefly unpack those five issues.

The Five Legal Guffaws

First, suppose a new AI law refers directly to purpose-built AI and omits stipulating that general-purpose AI is also encompassed by the regulation. For example, an AI law indicates that AI built for mental health guidance must meet stated restrictions. Do these same restrictions apply to a general-purpose AI that is also giving out mental health guidance?

Maybe, maybe not.

An AI maker of a general-purpose AI would surely argue that their AI is not within the scope of said law since the AI wasn’t built for that specific purpose. If the lawmakers had intended to cover all types of AI, they blundered in crafting the law to refer only to purpose-built AI. A glaring mistake by omission.

The same goes in the other direction. An AI law that calls out general-purpose AI and doesn’t mention purpose-built AI can end up letting that purpose-built AI escape the devised regulatory provisions. The builder of the purpose-built AI would insist that their AI is not a general-purpose AI. Thus, according to them, the law doesn’t apply.

There is also the mistake of writing a law that aims at one type and yet only names the other type. The law is missing the mark. It could be that a new AI law has all sorts of stipulations that apply to purpose-built AI. Unfortunately, the law only names general-purpose AI. The indicated stipulations might be askew and fail to cover the nuances of general-purpose AI.

Another form of blundering entails explicitly naming both purpose-built AI and general-purpose AI, but then lumping them together in a way that lacks sensibility. The crossover will bewilder the respective AI makers and builders. Clever legal arguments could be made to assert that the law is lacking in appropriate specificity and renders the law unusable.

Finally, it might be tempting to simply not say in a law whether it is intended for either type of AI. The hope of the lawmakers might be that the law covers both purpose-built AI and general-purpose AI. Or perhaps they hadn’t even considered the distinction while composing the law. This doesn’t solve the problem at hand. If anything, it gives an added opportunity for an AI maker or builder to contend that their type of AI wasn’t the type that the law was crafted to cover.

Lots of legal messes ensue, and society doesn’t benefit from whatever societal aims were in the hearts and minds of the lawmakers. For my 7-step common sense approach that lawmakers should consider when creating new AI laws, see the link here.

Latest Research On These Matters

Researchers are also exploring and exposing the troubles associated with AI laws that fail to suitably acknowledge these distinctions.

Historically, laws regarding AI were tilted toward purpose-built AI. Only recently has general-purpose AI (GPAI) become highly usable and prevalent. There has been a tidal wave of new AI laws that are arising due to the advent of GPAI. Do those prior AI laws on purpose-built AI apply to GPAI? Do the new GPAI laws overlap with the older laws? It is quite a legal stew.

In a notable paper entitled “Distinguishing Task-Specific and General-Purpose AI in Regulation” by Jennifer Wang, Andrew D. Selbst, Solon Barocas, Suresh Venkatasubramanian, arXiv, January 23, 2026, these salient points were made (excerpts):

“In recent years, the AI regulatory landscape has seen a chorus of concerns over AI safety and security risks.”
“Our goal in this work is to illustrate when the differences between general-purpose and task-specific AI are merely matters of degrees — where existing regulatory approaches remain applicable — and when they are differences in kind, demanding new regulatory instruments and forms of intervention.”
“Regulators should generally identify regulatory targets based on the applications and consequences of AI systems rather than technical implementation.”
“Identifying regulatory targets by technical specifications can still make sense as a last resort.”
“Regulators should focus on structural constraints and dependencies that prevent harm.”

The World We Are In

We need to straighten out the growing body of AI laws. This includes creating new AI laws that recognize prior AI laws and clarify where things stand. Lawmakers need to consider what the AI under the legal microscope does, the context of its usage, and the degree of control and responsibility for the AI. Do not sloppily allow underreach, nor overreach.

It is incontrovertible that we are now amid a grandiose worldwide experiment when it comes to society overall. The experiment is that AI is being made available nationally and globally, which is either overtly or insidiously acting to provide human guidance of one kind or another. Doing so either at no cost or at a minimal cost. It is available anywhere and at any time, 24/7. We are all the guinea pigs in this wanton experiment.

The reason this is especially tough to consider is that AI has a dual-use effect. Just as AI can be detrimental to humankind, it can also be a huge bolstering force for humanity. A delicate tradeoff must be mindfully managed. Prevent or mitigate the downsides, and meanwhile make the upsides as widely and readily available as possible.

Aristotle famously made this remark: “The law is reason, free from passion.” That makes abundant sense, but only if the law indeed contains well-devised logic and reasoning. New AI laws are being pumped out like sausages on an assembly line. Lawmakers must take a deep breath and ensure that their lawmaking covers the correct essentials of AI.

Source link