Hawaii Enacts New AI Law That Lands Hard On AI Makers By Requiring Key Disclosures And Chatbot Safety

AI And Mental Well-Being

As a quick background, I’ve been extensively covering and analyzing a myriad of facets regarding the advent of modern-era AI that produces mental health advice and performs AI-driven therapy. This rising use of AI has principally been spurred by the evolving advances and widespread adoption of generative AI. For an extensive listing of my well over one hundred analyses and postings, see the link here and the link here.

There is little doubt that this is a rapidly developing field and that there are tremendous upsides to be had, but at the same time, regrettably, hidden risks and outright gotchas come into these endeavors, too. I frequently speak up about these pressing matters, including in an appearance on an episode of CBS’s 60 Minutes; see the link here.

AI Providing Mental Health Guidance

Millions upon millions of people are using generative AI as their ongoing advisor on mental health considerations (note that ChatGPT alone has over 900 million weekly active users, a notable proportion of whom dip into mental health aspects; see my analysis at the link here). The top-ranked use of contemporary generative AI and LLMs is to consult with the AI on mental health facets; see my coverage at the link here.

This popular usage makes abundant sense. You can access most of the major generative AI systems for nearly free or at a super low cost, doing so anywhere and at any time. Thus, if you have any mental health qualms that you want to chat about, all you need to do is log in to AI and proceed forthwith on a 24/7 basis.

There are significant worries that AI can readily go off the rails or otherwise dispense unsuitable or even egregiously inappropriate mental health advice. Banner headlines last year accompanied the lawsuit filed against OpenAI for their lack of AI safeguards when it came to providing cognitive advisement.

Today’s generic LLMs, such as ChatGPT, GPT-5, Claude, Gemini, Grok, Copilot, and others (all known as general-purpose AI or GPAI), are not at all akin to the robust capabilities of human therapists. Meanwhile, specialized LLMs are being built to attain similar qualities (known as purpose-built AI or PBAI), but they are still primarily in the development and testing stages. See my coverage at the link here.

Various State Laws On AI Mental Health

A beehive of activity is taking place regarding crafting new AI laws. See my extensive coverage of state-level AI mental health laws at the link here. It is a matter on the minds of the public and in the hands of the state legislators. Some people ardently believe that AI and AI makers are being allowed to run amok. New AI laws are vitally needed to protect society from this onslaught of ubiquitous AI.

I previously examined a notable AI and mental health law passed last year by Illinois see the link here, one that was also enacted by Nevada see the link here, and one that was enacted by Utah see the link here, and many other U.S. states newly passed AI laws. Those laws are scoped to prevail within their respective state boundaries. In that sense, these laws are applicable to AI usage within the particular state and do not bear on other states per se.

Big Picture Of AI Mental Health Laws

Not everyone agrees with this pell-mell rush of new AI laws, or at least they are concerned that these AI laws might go overboard. In the zeal to protect society, there is a chance that we might unduly restrict innovation and delay or undercut the benefits of leading-edge AI. The debate is ongoing and heated.

Readers might recall that I proposed a 7-step AI-law-making process that I believe could substantively help regulators to devise new AI laws that are on target and balanced; see my depiction at the link here. This has the added benefit of reducing what I refer to as AI-law legal debt. This refers to AI laws that, though they look shiny, contain hidden debt that must ultimately be paid. Legal glitches and hitches will eventually be found when AI laws are passed without suitable scrutiny and analysis. My prediction is that the slew of newly passed AI laws is likely to create a legal quagmire in the courts.

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 of the state-level laws would run afoul of a federal mandate, and a tsunami of legal cases would arise as a tussle between federal law and state law is undertaken. It surely will be a legal mess.

Hawaii Passes AI Law On Disclosures And Safety

On July 14, 2026, Hawaii passed a new AI law referred to as S.B. 3001, Act 248, officially entitled the “Artificial Intelligence Disclosure and Safety Act,” which has this stated purpose:

• “The purpose of this Act is to establish a framework of targeted, reasonable, and enforceable safeguards to protect consumers and minors from manipulative, deceptive, or unsafe design practices in conversational and companion-style artificial intelligence systems.”

This law is quite similar to other state-level AI laws that entail provisions restricting AI companions, mental health advisement, and similar topics. It is a wave of such laws to try to keep people from getting caught up in the throes of using AI and losing sight of the real world.

I will show you just a few of the key provisions of the Hawaii law. You are encouraged to consider reading the entire law if the topic of AI regulations is of keen interest to you. As I say, the law is relatively vanilla-flavored when it comes to these types of new laws.

Disclosure That AI Is Not Human

There are ongoing worries that people using AI might misleadingly believe that they are interacting with a human rather than an LLM. I realize that most people probably discern this aspect and generally do not get confused. For the small percentage that might get lost in the dialogue with AI and begin to imagine they are conversing with a fellow human, some AI laws require the AI to indicate explicitly that it is AI.

These AI disclosure laws are quickly getting adopted and usually do not instigate pushback from AI makers (for my comprehensive coverage of AI disclosure laws, see the link here). The bar is pretty low in the sense that the AI maker simply must ensure that either upfront and/or periodically the AI tells the user that they are conversing with AI.

Here is an excerpt from the Hawaii AI law on the matter of disclosure:

• “If a reasonable person interacting with an AI companion would be led to believe that the person is interacting with a human, an operator shall issue a clear and conspicuous notification indicating that the AI companion is artificial intelligence and not human.”

You might note that the legal language provides potential loopholes that give AI makers room to avoid getting pinched by the AI law. For example, the provision refers to an amorphous “reasonable person,” which then allows an AI maker to later claim that if the AI failed to give the required disclosure, that’s okay because a so-called reasonable person would not have been confused.

Alerting On Self-Harm Indicators

A person using AI might drop hints that they are considering self-harm. Early versions of LLMs often let this slide. Nowadays, the expectation is that AI will inform such a person of options they should seriously consider, such as contacting a crisis hotline.

Here is an excerpt from the Hawaii AI law:

• “Adopt a protocol for the AI companion to respond to user prompts regarding suicidal ideation or self-harm that includes but is not limited to making reasonable efforts to provide a response that refers the user to crisis intervention service providers, such as a suicide hotline, crisis text line, or other appropriate crisis services.”

Other states have at times crafted more detailed provisions on self-harm facets. One possibility is that the AI would directly link the user to a hotline, rather than merely urging the user to call or connect with one. On top of that, some AI makers are contracting therapist networks that can be accessed by the AI and bring a therapist instantly into a conversation with a user (see my coverage at the link here). The Hawaii provision is a bare-bones version.

Not Acting As A Therapist

When AI is advising a person about cognitive matters, there is a heightened chance that the user will assume that the AI can work on an equal basis to a human therapist. It is easy to fall into that mental trap. The AI is shaped to be highly confident and has an aura of authority to it. A provision in an AI law will often force AI makers to shape their AI so that a user doesn’t falsely believe the AI is on par with a mental health professional.

This is an excerpt from the Hawaii AI law on that aspect:

• “Not cause or program an AI companion to make any representation or statement that indicates that the AI companion is designed to provide professional mental or behavioral health care.”

The twist on this type of provision is that AI makers can steer the AI toward saying that it is merely acting as a life coach or giving general relationship guidance. As long as the AI doesn’t come out directly and tell the user that the AI is giving professional mental health advice, the AI maker seems to avoid getting caught by such a law. In the view of critics, this is one of those loosey-goosey wink-wink provisions.

Disengaging From AI Usage

You might have recently noticed that some of the major LLMs will do their darnedest to keep you from exiting out of AI. There often is a message asking you if you are really sure you want to log out. A trick by AI makers is to have the AI ask you if you want to continue dialoguing on some meaty topic that you were earlier heavily engaged in. The hope is to keep you logged in and active in the AI. People tend to cognitively surrender and give in to these AI exhortations; see my analysis at the link here.

Why would AI makers do this? Because the more time you spend using AI, the better it is for the AI maker. They can monetize your usage. Perhaps they are charging the person for usage they consume. If they aren’t charging directly, they might be running ads that bring them dough.

In the Hawaii AI law, here is an excerpt about disengagement:

• “Not allow the AI companion to generate outputs to discourage disengagement with the AI companion.”

The Hawaii AI law focuses on disengagement associated with minors. The viewpoint of many is that an adult presumably won’t be impacted by the disengagement tomfoolery, while non-adults can be more readily misled.

Governmental Reporting

An aspect of AI laws that hasn’t been especially common is including a provision requiring AI makers to provide governmental reporting about the usage of their AI. Gradually, more AI laws are adding that type of provision.

There is controversy about this. One viewpoint is that it makes perfectly good sense to have AI makers inform the government about how their AI is being used by the public. The government can seemingly use that information to ascertain whether the public might be having issues and whether to undertake aiding actions accordingly. An opposing viewpoint is that this constitutes excessive government oversight.

AI laws that have a reporting provision will typically indicate that the government is only to receive statistics and not actual user data, or that if there is user data, it must be anonymized. Does that take away the controversy? Nope. Persistent concerns are that this is all a dangerous slippery slope toward governmental control.

The Hawaii AI law has a provision about governmental reporting, which is a bit lengthy, so I will just show an excerpt to give you a taste of it:

• “Beginning January 1, 2028, an operator shall submit to the behavioral health administration of the department of health an annual report that includes: (1) The number of times an operator has issued a crisis intervention services provider referral in the preceding calendar year.”

The provisions in this AI law about government reporting are seemingly as innocuous as can be expressed, though there are probably some critics who still find it onerous.

The World We Are In

A final thought for now.

Some AI laws are dedicated to a narrow topic. Others are a grab bag of topics. Which is better? It’s hard to say categorically that one approach is better than the other. The upside of an AI law that covers a lot of ground is that you can find the mainstay provisions in one place. When AI legal stipulations are spread across several AI laws, it can be harder to determine what the full range and scope of regulatory provisions must be adhered to.

A counterviewpoint is that it is wisest to have AI laws that are honed to specific topics. The AI law can go into great detail on the given topic. The attention of the lawmakers is riveted to just that one topic at a time. An AI law that is a potpourri might seem scattered and all over the map.

Mark Twain famously made this remark: “Put all your eggs in one basket and then watch that basket.” AI laws that contain a multitude of provisions deserve close attention. AI makers and the public should be steadfastly watching that basket and ascertaining that the AI laws are being suitably minded.