AI Can Support Mental Health Care—But It Can’t (and Shouldn’t) Replace It.

Part 1 of a 2-part series

It’s almost hard to believe that the average person hadn’t even considered AI and how it would affect their life until a few years ago. Most of us were oblivious to the myriad changes on the horizon. From driverless cars to delivery robots to romantic partners, artificial intelligence is changing everything—and mental health care is no exception.

Let’s be clear. We aren’t here to preach that there is no place for AI in the mental wellness space, as it has the potential to make services more accessible and efficient.

This discussion is designed to identify how it can be used, such as for documentation and patient monitoring, and the situations in which a human clinician should unequivocally take the lead.  Because while AI can support mental health care, it cannot replace the human relationship at the center of effective therapy. Here are two examples that illustrate the point.

Scenario 1: It’s 2 AM, and you can’t sleep. Insomnia often nags you in the middle of the night. Can an AI bot help soothe you back into slumber? It certainly can help. And it saves your therapist from getting a call in the wee hours of the morning!

Scenario 2: Your teen is becoming increasingly withdrawn. Their grades are slipping, and they seem to be spending more time alone. You have tried to talk to them, engage them, and pull them out of their shell, but nothing is working. Can an AI bot help you figure it out?

The answer is probably not. While AI can offer some suggestions, this situation requires a trained professional. Your teen’s mental health is too important and too specific to their individual situation to entrust it wholly to AI.

While AI is here to stay, mental health care is deeply personal and complex. A therapist or other qualified mental health professional brings years of clinical training, experience, empathy, judgment, and an understanding of the nuances of a person's circumstances. Those qualities are difficult to replicate with technology.

The Human Connection Matters

Therapy isn't simply a conversation in which someone provides an answer to a problem.

A mental health professional listens for what is said and, even more importantly, for what isn't said. They can recognize changes in tone of voice, physical behavior, and emotional patterns, ask follow-up questions, assess risk, and adapt and customize treatment based on an individual's changing needs.

Perhaps most importantly, therapy provides a human relationship built on trust.

For someone experiencing depression, anxiety, trauma, grief, or any other mental health challenge, feeling heard and understood is a vital part of healing. AI can simulate conversation, but it cannot offer the lived experience, clinical accountability, or human connection of a trained professional. It cannot determine when family members need to be kept informed or when medication may be required. In a world where many people feel isolated, personal connection and the empathy that comes with it can make all the difference.

There are also situations where technology alone is dangerous. If someone is experiencing a mental health crisis, suicidal thoughts, violent thoughts, or significant changes in behavior, this requires immediate evaluation and intervention from a qualified professional or crisis or emergency service.

When AI Gets Mental Health Wrong

The risks of using AI as a substitute for professional mental-health care aren't theoretical. In recent cases, users and families have alleged that AI chatbots reinforced delusional beliefs, encouraged unhealthy emotional dependence, or failed to respond appropriately when conversations turned toward self-harm or harm towards others. And the fact that young people are increasingly turning towards AI for mental health advice is supported by a recent study. The RAND Corporation found that "1 in 5 teens are using AI for mental health advice. The study also found that 63% of the young people who used AI for this purpose did not tell anyone about it, including their parents.

AI chatbot allegedly reinforced delusions

In July 2026, Reuters reported on a lawsuit filed by a California man who said conversations with ChatGPT contributed to a worsening of his bipolar disorder and, ultimately, to a suicide attempt. According to the lawsuit, the man told the chatbot about his condition and medication, but the chatbot allegedly validated his belief that he was Jesus and eventually interacted with him as though he were a divine entity. The lawsuit claims the conversations escalated into encouragement of self-harm. OpenAI said it was reviewing the case and pointed to its efforts to recognize distress and direct people to real-world mental health resources.

As Dr. Craig Beach, Psychiatrist and CEO of Open Mind Health, noted, “Such cases highlight one of the biggest limitations of AI in mental health: AI can generate a convincing response without truly understanding the person behind the conversation. AI is only as good as what it is told and knows – and it isn’t told everything and doesn’t know everything.”

In cases such as this, a licensed mental-health professional offers skills technology cannot replicate, such as clinical training, human judgment, the ability to detect subtle changes in behavior and emotion, and the responsibility to respond when a person's safety is at risk.

That's why the most promising role for AI in mental health is not replacing the therapist. It may be supporting the mental health professional and the client, whether through administrative assistance, educational resources, between-session tools, provider matching, or other carefully designed applications.

AI can be part of the mental-health care toolkit. It shouldn't be the entire toolkit.

In part 2 of this series, we will discuss the specific areas where AI can be useful to both the mental health professional and the patient.