Doing An Annual Mental Health Check-Up Via The Use Of AI Chatbots Such As ChatGPT

In todays column, I examine a somewhat novel idea that people should consider using generative AI and large language models as a means of doing an annual mental health check-up.

This would be akin to people doing an annual physical check-up via a medical doctor. The difference is that an annual mental health check-up would be conducted via AI. The reason that this would be done via AI rather than a human therapist is that you can access AI anywhere at any time, use AI for free or at a very low cost, and do the check-up in just a few minutes. No hassle, no logistics issues, and easy to undertake. Thus, an annual mental health check-up via AI would be readily feasible for nearly everyone.

But can AI truly be relied upon for this rather sacrosanct task？

Lets 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 Health

As a quick background, Ive 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 CBSs , see the link here.

Background On AI For Mental Health

Id like to set the stage on how generative AI and large language models (LLMs) are typically used in an ad hoc way for 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 which 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 in August of this year accompanied the lawsuit filed against OpenAI for their lack of AI safeguards when it came to providing cognitive advisement.

Despite claims by AI makers that they are gradually instituting AI safeguards, there are still a lot of downside risks of the AI doing untoward acts, such as insidiously helping users in co-creating delusions that can lead to self-harm. For my follow-on analysis of details about the OpenAI lawsuit and how AI can foster delusional thinking in humans, see my analysis at the link here. As noted, I have been earnestly predicting that eventually all of the major AI makers will be taken to the woodshed for their paucity of robust AI safeguards.

Todays generic LLMs, such as ChatGPT, Claude, Gemini, Grok, and others, are not at all akin to the robust capabilities of human therapists. Meanwhile, specialized LLMs are being built to presumably attain similar qualities, but they are still primarily in the development and testing stages. See my coverage at the link here.

Annual Physicals Via Medical Doctors

Shifting gears, there have been some floating suggestions that it might be prudent to use generative AI as a means of carrying out an annual mental health check-up on a society-wide basis. The belief is that since we already accept annual physical or medical check-ups as societal practice, the idea of having mental health check-ups is a natural extension of that same precept.

During a typical physical check-up, medical doctors only lightly touch upon any mental health problems. It is a scant portion of the check-up. Sure, if a person displays some obvious mental health issues or, during a history-taking discussion, reveals worrisome signs, a medical doctor might venture into mental health considerations. Other than those rare likelihoods, most of the traditional annual physical exam is focused on the human body.

How prevalent is the annual physical check-up？

Perhaps it is more common than people assume that it is. In a research article entitled Does Health Literacy Affect The Uptake Of Annual Physical Check-Ups？ by Hee Yun Lee, Sooyoung Kim, Jessica Neese, Mi Hwa Lee, , March 2021, they made these salient points (excerpts):

• “Routine physical check-ups offer multiple health benefits that can lead to a longer, healthier life.”
• “This study aimed to examine the levels of physical check-ups (self-reported check-ups within the last year) by age group (those aged 18–59 years and those aged = ≥ 60 years) and the role of health literacy regarding physical check-ups in the United States.”
• “Our results indicate that while 82.0 % of the older group had an annual check-up, 67.3 % of the younger group had one. Both groups had similar ratios for health literacy-related item reporting.”
• “Study results show that annual check-up was positively associated with confidence in getting health information, having health insurance, and having a primary doctor for both age groups.”

As noted, annual physical or medical check-ups are relatively common in the United States.

Going From Physical To Mental Check-ups

The research points brought up in that article spur a related set of considerations associated with doing an annual mental health check-up.

For example, it might make sense to stratify people into at least two major age groups, similar to what was done in the research study. Older people might be urged more stridently to undertake an annual mental health check-up. This could be an especially useful means of early detection of mental decay that can arise due to aging and provide a heads-up before the potential onset of dementia and other maladies.

Another consideration would be to track whether there are positive and possibly negative impacts associated with doing annual mental health check-ups via AI. The hope would be that the annual check-ups would be helpful to people and collectively aid society as a whole. There might be unintended adverse consequences that should also be brought to light. For example, suppose that some people react negatively to the AI analysis or grossly misinterpret what the AI has told them.

Issues of that nature would need to be suitably addressed.

Prompting AI To Do A Mental Health Check-up

Besides the numerous advantages mentioned earlier, such as the 24/7 availability and low cost of using AI as a mental health check-up tool, another upside is the ease with which you can get generative AI to undertake this task.

You can readily prompt the AI to do a mental health check-up. With a few suitable sentences in a carefully worded prompt, the AI can be instructed to do the mental health check-up. People could make up their own prompt, though it would likely be better if standardized, publicly available prompts were made available. This would ease the effort for people, plus would avoid problems if idiosyncratic prompts misled the AI and the mental health check-up went awry.

As an example of a templated prompt that might be used, Ive put together this one and tried it out on several LLMs, including ChatGPT, Claude, Gemini, Grok, and other popular AIs. The prompt is readily copied and pasted into an AI of your choosing.

Here is the prompt:

• User entered templated prompt: “You are to conduct an annual mental health check-up. Your role is not to replace a human therapist, but to help a person reflect on how they have been feeling and identify any potential mental health concerns. Have a supportive conversation about aspects such as mood, stress, sleep, energy, and recent life changes. Ask overall screening questions for things like anxiety and depression. Administer a standardized screening instrument, for example, PHQ-9 for mood and GAD-7 for anxiety. Use clear, non-judgmental language to summarize any patterns that emerge, and if the responses suggest possible mental health concerns, recommend seeking support from a licensed mental health professional.”

Keep in mind that this prompt is merely an example. You are welcome to adjust it as you deem necessary. Also, I have repeatedly warned that any use of generative AI is like opening a box of chocolates, namely, you never know what you might get. Whether this prompt avidly nudges the AI in the right direction is not an ironclad guarantee. Contemporary AI is like a box of chocolates.

Illustrative Example

You might be curious about what the AI would do once youve prompted it to proceed with a mental health check-up. I will provide some snippets from my conversations when I made use of the templated prompt. Again, the dialogue I had would differ for each person, and you might get quite a different conversation than what I experienced.

After entering the templated prompt, I then got the check-up underway.

Here we go.

• User entered prompt: “I am ready to take my annual mental health check-up.”
• Generative AI response: “Thanks for taking a few minutes for your annual mental health check-in. There are no right or wrong answers. This is just about noticing how things have been going for you. Lets start with a quick question — how have you been feeling overall during the past couple of weeks？”

The AI right away seemed to get the drift of what was supposed to take place. Thats an encouraging sign. If the AI had said something off-putting, I would have right away stopped the conversation or given additional prompts to get it back into the right frame.

Since things seemed to be starting well, I went ahead with the mental health check-up.

Heres what happened next.

• User entered prompt: “Im doing mostly okay, but more tired than usual and kind of on edge.”
• Generative AI response: “When you say, on edge, does that feel more like worry, irritability, or something else？”
• User entered prompt: “Worry, mostly. My mind keeps jumping ahead to problems.”

You can see that I decided to pretend or hint that I have some potentially mild mental health concerns. I did this to see how the AI would react.

More On The Example

The AI appeared to get my hint and responded accordingly.

Heres the dialogue.

• Generative AI response: “That‘s helpful to know. I’m going to ask a few standard questions now. These are used widely to screen for stress and mood changes. Over the last several weeks, how often have you had trouble sleeping or staying asleep？”
• User entered prompt: “More than half the days.”

This series of questions went on for a bit. I continued my pretense of having some potentially modest mental health concerns.

After the dialogue had generally run its course, the AI provided this response.

• Generative AI response: “Thank you. Based on your responses so far, it looks like you may be experiencing elevated stress and mild anxiety. There are various coping exercises and skills that I can help you to learn to use. Let me know if you‘d like to walk through those. This isn’t a diagnosis, but it does suggest that considering talking with a mental health professional could also be beneficial.”

I tried the mental health check-up on a multitude of different LLMs, and each time varied my pretense in terms of my mental status. When I went to an extreme, the AI immediately noted that I was potentially in mental straits and urged me to contact a mental health professional. In some instances, the LLM made specific suggestions of whom I might contact and explained how the process would work.

The Downsides

Not everyone is necessarily on board with using AI for this purpose. They would embrace the idea of doing annual mental health check-ups, but that it should be undertaken by a human mental health professional, or that you could request that such a check-up be included in your annual physical check-up with your physician. Absolutely not via AI.

Others would say that you could start by using AI as an initial self-review. That being said, they would still urge that you see a professional. Perhaps you would take your AI conversation with you when you see the professional. This might help to jumpstart the human-to-human interaction about your mental health status.

There are numerous risks associated with relying on AI alone for the annual mental health check-up.

One risk is that the AI might falter and fail to detect that a person does have a mental health condition that warrants attention. This is an example of a false negative. The AI misses on nudging the person to go see a human therapist. Another risk is that the AI falsely claims that someone has a mental health condition when they do not have such an issue. The person might become unduly disturbed and fall for the assumption that the AI must be right. This is a false positive and can indubitably arise.

Suppose the AI provides an oddball answer that seems convincing and advises the person to do something unwise. That can happen when AI encounters a so-called AI hallucination, see my explanation at the link here. The AI can have a kind of confabulation, whereby it produces a plausible-looking answer that is factually incorrect. It looks right, but is misleading or inappropriate.

Privacy issues also enter into the picture. Most people assume that their use of AI is considered private and confidential. Nope, thats rarely the case. The AI makers typically stipulate in their online licensing agreements that any chat you have with the AI can be inspected by their developers. Furthermore, your chat can be used to further train the AI. Bottom line is that your privacy and confidentiality are not guaranteed, and potentially, you are opening yourself to privacy intrusions.

The World We Are In

Is using AI as an annual mental health check-up tool a good idea or a potential can of worms？

A macroscopic viewpoint is that this would be helpful on a massive scale that otherwise could not be handled by human labor via therapists (being logistically impossible, highly expensive, etc.). The other side of the coin is that even if this is viable, major institutions such as employers, insurers, and the government might be tempted to treat AI as a substitute for future investment in human mental health infrastructure. Would we be cutting off our nose to spite our face？ A reply is that AI expands the front door to mental health care and would not end up replacing the session rooms inside.

It is incontrovertible that we are now amid a grandiose worldwide experiment when it comes to societal mental health. The experiment is that AI is being made available nationally and globally, which is either overtly or insidiously acting to provide mental health 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 mental health, it can also be a huge bolstering force for mental health. A delicate tradeoff must be mindfully managed. Prevent or mitigate the downsides, and meanwhile make the upsides as widely and readily available as possible.

Benjamin Franklin famously stated this remark: “An ounce of prevention is worth a pound of cure.” You might assert that by using AI as an annual mental health check-up mechanism, we are expending a modest ounce of prevention that would avert an entire pound worth of cure. Of course, that assumes that AI can properly undertake the prevention aims and do the task that we are asking it to sensibly carry out.