Vermont Passes Chatbot Mental Health Law That Raises Questions About Therapists Rubber-Stamping AI Advice

New AI mental health law by Vermont raises the question of therapists possibly rubber-stamping AI.

getty

In today’s column, I examine a new law in Vermont regarding the use of AI for mental health advisement. Overall, the law is generally on par with many other states that have enacted similar laws. There aren’t any especially unusual conditions or extraordinary requirements.

That being said, the Vermont law is a handy vehicle for considering a frequent and potentially challenging stipulation in these budding state-level laws about the oversight of AI for mental health. The key indication in most of these laws is that a therapist or mental health professional must be somewhere “in the loop” and serve as a reviewer and potential approver of any mental health recommendations or similar outputs from the AI.

On the face of things, this makes sense. A professional therapist will presumably catch the AI in providing untoward or misguided mental health advice to users. One worry is that therapists performing this oversight role will potentially fall into a classic trap, namely, they will begin rubber-stamping whatever the AI says. This is quite possible and usually due to the therapist being lulled into believing that the AI is always right, plus the therapist might be too busy to fully review what the AI has said, and otherwise, a therapist might do a cursory assessment for a variety of reasons. Some assert that these new AI laws need to put much stronger teeth into the requirements so that therapists will stay fully diligent and remain on their toes when it comes to AI chats.

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 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. 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.

Vermont Passes AI Mental Health Law

Vermont passed a new AI mental health law, H.816 (Act 156), titled “An act relating to regulating the use of artificial intelligence in the provision of mental health services.” The legislation was signed into law on June 17, 2026.

This law is quite similar to most other state-level AI mental health laws. The regulation stipulates that AI makers aren’t to provide AI in Vermont that performs mental health advisement unless the AI is overseen by a professional therapist. It is a wave of such laws to try to keep people from getting caught up in unfettered AI advisement that improperly or inappropriately misguides them cognitively.

I will show you just a few of the key provisions of the Vermont law. You are encouraged to consider reading the entire law if the topic of AI mental health 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.

What The Law Stipulates

Here are some key excerpts from Vermont’s H.816 (Act 156):

• “It is the purpose of this act to safeguard individuals seeking mental health services in Vermont from psychological harm, including death by suicide, by ensuring that these services are delivered by mental health professionals and not independently by artificial intelligence systems.”
• “A corporation or entity shall not provide, advertise, or otherwise offer mental health services, including through the use of artificial intelligence, to the public unless the mental health services are: (1) provided by a mental health professional.”
• “Nothing in this section shall preclude a mental health professional who is operating within the professional’s scope of practice from utilizing artificial intelligence tools that are compliant with the Health Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191, provided that the mental health professional reviews and approves any mental health services.”

The emphasis is that a mental health professional must review and approve the AI mental health services.

One aspect that I’d like to bring to your attention is that this law fortunately does not seem to require that the therapist be logged in and actively monitoring the AI during a mental health chat with someone. I mention this because a new law in Colorado requires that therapists access AI on a real-time, synchronous basis so that they can instantly intervene in a user chat if needed. In my view, Colorado’s law goes too far, as explained in detail in my analysis at the link here.

Therapists Can Be Lulled Into Complacency

Imagine that a therapist becomes familiar with a specific AI that they have agreed to monitor on an oversight basis. Suppose each time they double-check the AI, it has done a satisfactory job of providing mental health guidance. This happens repeatedly. Eventually, the therapist is going to assume that the AI is likely passable and that the need to closely double-check the AI is no longer needed.

The therapist starts to pull back on how much effort and diligence they undertake when reviewing the AI. It’s a natural reaction and can happen in nearly any line of work. Sometimes this is referred to as automation bias. You get comfortable that the AI is going to do the right thing.

The problem is that the therapist might miss situations that involve the AI going off the deep end or otherwise dispensing unsavory mental health advice. Some instances could readily slip between the cracks. Consider some numbers. Of perhaps five hundred AI mental health recommendations they have reviewed, maybe only a few are wrong, and the therapist didn’t catch them. Let’s pretend it is just 5 instances. Meanwhile, the therapist has gotten into a rubber-stamping mode and approved all 500 chats.

You might argue that the therapist is only faltering by a tiny percentage, maybe 1% or less; so, it’s no big deal. A contrary viewpoint is that this is a horrendous track record since there will be five people who are potentially abiding by adverse mental health advice produced by the AI – and a human therapist signed off on it.

Laws Specifying What Review And Approval Consist Of

By and large, these new AI laws do not clarify in any exacting way the nature of the review and approval that is envisioned by the spirit and intention of the regulation. The act of reviewing is left unspecified. The approval elements are also unspecified.

An approach could be to offer sufficient detail in these laws to lay out what a therapist is truly obligated to do. If a therapist falters and they end up in court, the courts will likely proceed to examine what semblance of review occurred and how the approval was recorded and communicated. We can anticipate this and instead load those needed aspects into the law itself.

Consider these types of questions that would customarily be asked:

• Did the therapist actually read the AI mental health recommendations?
• Did the therapist independently evaluate the clinical reasoning?
• How much time was spent reviewing?
• Was documentation created showing the review that occurred?
• When did the therapist subsequently approve the AI recommendations?
• What was the logical basis for the approval by the therapist?
• Did the therapist create documentation identifying the approval?
• Could another clinician reconstruct the review process?

Without these kinds of statutory standards, a "review" might be wholly cursory and done in the blink of an eye. The approval might equally be undertaken without careful consideration. Therapists can become complacent. The law ought to provide a sufficient means to overcome or prevent that complacency from arising.

There’s an added twist. Some therapists might use AI to do the review and therefore never directly inspect the AI mental health advice at all. Without credible and specified requirements for human review and approval, therapists are going to fall into the trap of allowing the AI to review and approve itself. Not good.

The Back And Forth On Duty

There are tradeoffs involved in this. Some might clamor that therapists are professionals and that these new AI laws should not overtake or insult their professional integrity. The argument is that they know what must be done and there isn’t a need to get too much into the weeds. Let the professional handle things.

A rejoinder is that not all therapists will necessarily know how to handle managing AI in the role of a human reviewer and approver. For many therapists, AI is new to them. They might not realize they do need to stay on top of what the AI indicates. They might not be aware of automation bias as a form of slippery slope. By baking the review and approval aspects into a law, this at least provides an upfront notification.

Therapists might also not sense that they are taking on a heavy responsibility in this AI oversight role. Licensing boards haven’t actively stepped into these matters because these laws are only now emerging. Once licensing authorities do intercede, the odds are that therapists will begin to see a range of sanctions being imposed on those who let down their guard on AI. This could include formal reprimands, probation of their therapy license, suspension, revocation, and the like.

The sword dangling over the heads of therapists has many sharp edges. You can bet that malpractice lawsuits will be formulated and bolstered based on these new AI laws. Courts might also view these failings as evidence of negligence on the part of a therapist. For those therapists who go the disconcerting route of merely clicking on approval and not undertaking any review at all, there is a chance of allegations of fraud or false certification coming into the mix.

Diffusion Of Responsibility

I’ve got a mind-bending final consideration on this that you might want to solemnly ponder.

Due to these new AI laws, AI makers might point fingers at therapists and proclaim that even if the AI gave out bad mental health advice, the AI maker shouldn’t bear the responsibility since the reviewing therapists should have found it. Bam, drop the mic. But then therapists might declare that if they did happen to overlook when AI generated bad advice, this is entirely on the backs of the AI makers since the AI maker should have ensured that the AI wouldn’t do so. Period, end of story.

There is a chance that this finger-pointing and diffusion of responsibility will give both sides excessive wiggle room. Neither will admit that the buck stops with them. It’s an excuse-making opportunity, while to the rest of the world, the buck stops with both.

As per the wise words of Abraham Lincoln: “You cannot escape the responsibility of tomorrow by evading it today.” Let’s craft AI laws on mental health that are sensible, complete, and will deliver uplifting results.