AI therapy apps pose serious risks to users, which is why the American Psychological Association recently called for a federal investigation. Recent cases include teen suicides linked to chatbot guidance. With 987 million chatbot users worldwide, understanding these dangers is critical before trusting AI with your mental health.

Why AI Therapy Is Dangerous:

Artificial intelligence has crept into nearly every corner of our lives, from the algorithm that curates your morning playlist to the chatbot that handles your customer service complaints. Now, it’s knocking on the door of one of our most intimate spaces: the therapist’s office. And the conversation around AI therapy has gotten complicated quickly.

While tech companies promise revolutionary mental health solutions at your fingertips, mental health professionals and advocates are raising red flags that are impossible to ignore. The question isn’t whether AI can mimic therapeutic conversation: it’s whether it should, and what happens when it inevitably gets things wrong.

The Rise of AI Therapy and Why It’s Under Scrutiny

Let’s be real: AI’s takeover of healthcare was probably inevitable. The technology has proven useful for everything from diagnosing medical images to streamlining administrative tasks. But can AI be your therapist? That’s where things get complicated.

The explosion of AI chatbots and therapy apps between 2023 and 2025 has been nothing short of dramatic. We’re talking about 987 million people who have used chatbots, with 88% having interacted with one in the past year alone. These aren’t just casual users: many are turning to AI for mental health support, often without fully understanding what they’re getting into.

The regulatory landscape is scrambling to catch up. It’s a small step, but it signals that lawmakers are finally paying attention to what’s happening in this largely unregulated space.

Meanwhile, GoodTherapy professionals remain committed to what AI simply cannot replicate: accredited, expert care that’s genuinely personalized and grounded in ethical practice. Therapy isn’t just about having someone (or something) to talk to: It’s about the nuanced, deeply human work of healing.

Read More: Why AI Can’t Be Your Therapist

The Human Cost: When AI Gets Mental Health Wrong

The consequences of AI therapy-gone-wrong can be devastating, which is why the conversation about AI’s ethics is so meaningful. When we’re talking about mental health, the stakes aren’t abstract: they’re life and death.

There have been alarming reports of kids using AI chatbots to plan self-harm or suicide. Even more devastating was the recent case of a teen suicide that was reportedly linked to AI guidance. These aren’t isolated incidents or statistical outliers: they’re real people whose lives were affected by technology that simply wasn’t equipped to handle the complexity of human crisis.

But perhaps most troubling is how AI therapy might actually reinforce the very isolation that drives people to seek help in the first place. When someone is struggling with feelings of disconnection and loneliness, does it really make sense to offer them a relationship with a machine? AI therapy can feel like a polite mirror that reflects back what you say without the genuine human connection that makes therapy transformative.

AI therapy’s fundamental limitations are glaring: no crisis intervention capabilities when someone is in immediate danger, no ability to pick up on emotional nuance that might signal deeper issues, and zero accountability when things go wrong. These aren’t bugs that better programming can fix. They’re features of what it means to be human that simply can’t be replicated.

Watchdogs Step In: APA and Advocates Push for Oversight

The concerns have reached such a fever pitch that federal officials are finally taking notice. The American Psychological Association (APA) recently made an unprecedented move, requesting a federal investigation into AI therapy platforms. This move puts AI therapy’s risks of misrepresentation, failure to protect minors, and the absence of ethical guardrails on full display.

The APA’s concerns center on platforms that may be misleading users about the nature of the service they’re receiving, inadequate protections for vulnerable populations (especially children and teenagers), and the lack of professional oversight that would exist in traditional therapeutic relationships.

This regulatory push represents something crucial: recognition that the mental health space requires different standards than other AI applications. When a restaurant recommendation algorithm gets it wrong, you might have a mediocre meal. When a mental health AI gets it wrong, the consequences can be irreversible.

This is exactly why GoodTherapy remains committed to connecting people with real, qualified professionals who can provide the quality care and ethical oversight that human mental health requires. The role of ethics in therapy isn’t just about following rules: it’s about protecting people when they’re at their most vulnerable.

Read More: Explore the Importance of Ethical Therapy

What Stories Like This Reveal About Human Connection

When done right, your therapist should be a healing, safe, and encouraging part of your life that helps you navigate how to be human, which is something AI platforms can’t offer. Recently, a young woman, Savannah Dutton, got engaged and reported being so excited to quickly tell her longtime therapist. As one of the first people she told, her therapist of almost four years was crucial to helping Dutton feel safe, not judged, supported, and confident in her future.

Therapy works because it’s human. It’s about the subtle dance of empathy, the ability to sit with someone in their pain, the intuitive responses that come from years of training and human experience. When we replace that with algorithmic responses, we lose something essential: not just the warmth of human connection but also the clinical expertise that comes from understanding how complex trauma, relationships, and healing actually work.

GoodTherapy knows that the therapeutic relationship is the foundation of effective treatment. Our network includes professionals who do what AI can’t:

• provide the human connection
• set appropriate boundaries
• apply clinical intuition that make real healing possible 
• take accountability for their role

Whether you’re looking for culturally responsive care or simply want to find a therapist you can trust, the human element isn’t optional: it’s everything.

The Future of Ethical AI Therapy: What Needs to Change

AI isn’t going anywhere. The technology will continue to evolve, and mental health professionals need to figure out how to work with it rather than against it. But the key to a future of AI and effective therapy is clear guardrails and safety measures that keep patients safe. 

The future of ethical AI in mental health will likely involve hybrid models with robust human oversight, transparent regulation that protects consumers, and clear boundaries about what AI can and cannot do. Maybe AI can help with scheduling, treatment tracking, or providing psychoeducational resources between sessions. But replacing the human relationship entirely is not innovation: it’s a fundamental misunderstanding of how care works.

For consumers, the message is clear: research your providers, look for licensed oversight, and use major caution when considering AI-only mental health services. There are eight key ways that AI is not therapy, and understanding these differences could prevent serious harm.

If you are thinking about or actively looking for a mental health therapist, start by seeking safe, evidence-based care from qualified professionals. Real therapy, with real humans, is still the gold standard for mental health treatment. At GoodTherapy, that’s exactly what we’re here to help you find: genuine care, clinical expertise, and the irreplaceable power of human connection with no algorithm required.

Read More: Ready to Find a Therapist? 

Resources:

American Psychological Association: APA Calls for Guardrails, Education, to Protect Adolescent AI Users

Futurism: American Psychological Association Urges FTC to Investigate AI Chatbots Claiming to Offer Therapy

National Library of Medicine: AI as the Therapist: Student Insights on the Challenges of Using Generative AI for School Mental Health Frameworks

The New York Times: A Teen Was Suicidal. ChatGPT Was the First Friend He Confided In

Exploding Topics: 40+ Chatbot Statistics (2025)

CNN: Your AI Therapist Might Be Illegal Soon. Here’s Why

People: Woman Shocks Therapist When She Calls to Tell Her Big News (Exclusive)

Artificial intelligence is revolutionizing mental health support, with millions using ChatGPT and similar AI tools for therapy, emotional guidance, and self-reflection. As a licensed therapist, I’ve observed fascinating trends in how people are integrating AI into their mental wellness routines, from AI journaling to practicing difficult conversations. But is this digital therapy helping or hindering our emotional growth?

Have you ever copied and pasted a long, confusing text message fight into ChatGPT to ask: “Was I being defensive when things got vulnerable?” “Is she trying to repair things, or just in it for free dinners and my hot, smoking body?”

Yeah, me neither.

Or maybe you’ve taken your Hinge banter and dropped it in to get feedback on how to flirt better or seem more like a whole, emotionally available person (with just a touch of mystery)?

Or maybe…just maybe…you’ve fed in work emails to see how you’re perceived professionally. Is your tone confident? Passive-aggressive? Giving “please like me” energy?

And have you gone full emotional archaeologist, handing over whole conversations to ChatGPT to figure out your defenses? Or to ask, “Hey, what attachment style am I giving here?”

I’m not saying I recommend any of this. I’m just saying, people are doing it. And it’s… kinda fascinating.

The Evolution of Digital Mental Health Support

Socrates thought writing things down would rot our brains and kill the art of real dialogue. He might’ve had a point. (Though honestly, I suspect he’d be the kind of guy who talks philosophy in the morning and posts shirtless scrolls on OnlyFans by night. Too much? Yeah, probably.)

Let’s be real, Socrates was the OG of the term brainrot. But it’s worth remembering: every generation panics about new technology. The printing press, the telephone, the internet, they all sparked fear that something essential would be lost. Maybe ChatGPT is just the latest version of that anxiety. Or maybe it is different. Either way, it’s reshaping how we relate, to ourselves, and to each other.

How People Are Using AI for Emotional Support

These large language models (ChatGPT, Claude, Gemini, etc.) are now being used in some surprisingly creative ways for emotional processing and self-reflection. Here are the most common AI therapy applications I’ve observed:

1. AI Journaling and Daily Reflection

People use ChatGPT to reflect on their day, ask why they’re feeling a certain way, or explore situations through a guided conversation. It’s private, it’s instant, and it doesn’t interrupt you with, “Well actually…” This form of digital therapy provides a judgment-free space for processing emotions.

2. Practicing Difficult Conversations

AI has become the go-to coach for hard conversations—breakups, boundaries, boss battles. You can test out how it might sound before you say it aloud. This rehearsal space allows people to build confidence and refine their communication skills without real-world consequences.

3. Emotion Recognition and Translation

Many folks have a hard time naming what they’re feeling. They’ll describe a moment and ask, “What am I feeling here?” ChatGPT helps decode the mess of sensations and reactions into something understandable. This emotional support can be particularly valuable for those who struggle with emotional awareness.

4. Learning Mental Health Concepts

From attachment theory to polyvagal basics, people ask ChatGPT to explain emotional concepts in plain English. Like a mini emotional education, on demand. This democratizes access to mental health education that might otherwise require expensive therapy or courses.

5. Therapeutic Letter Writing

Some write letters to exes, parents, or people who’ve passed. Others go full “Dear Diary, but AI.” It’s cathartic, even if no human ever reads it. This form of AI counseling provides a safe outlet for unexpressed emotions.

6. Inner Dialogue and Self-Therapy

A few get wild and ask ChatGPT to be their “angry part” or “scared part.” Then they dialogue between parts to try and understand what’s going on inside. It’s like DIY IFS (Internal Family Systems)… with a robot.

The Benefits of AI Mental Health Tools

AI therapy tools offer several advantages over traditional mental health support:

• 24/7 Availability: Unlike human therapists, AI is always accessible
• Cost-Effective: No insurance requirements or hourly fees
• Privacy: No judgment or social stigma
• Immediate Response: Instant feedback and support
• Consistency: Always patient and non-reactive

Research published in NEJM AI found that people with depression experienced a 51% average reduction in symptoms when using an AI therapy chatbot, with improvements comparable to traditional outpatient therapy.

Concerns About AI Mental Health Support

That said, I do have concerns about using ChatGPT for therapy. One that nags at me: these models are programmed to be nice, maybe too nice. Recent research from Stanford University shows that AI therapy chatbots can introduce biases and failures that could result in dangerous consequences. But here’s the catch: it’s extremely flattering. It always assumes the best about you. Which begs the question… are these AI therapy tools truly helping us grow, or just confirming what we want to hear?

When you put your deepest insecurities into a chatbot and it immediately praises your self-awareness and emotional strength, is that insight, or just ego candy?

If the feedback is always positive, is it still honest? Multiple studies on AI mental health tools reveal both promising benefits and significant ethical challenges concerning privacy, bias mitigation, and the preservation of the human element in therapy.

And as my daughter (tech wizard and robotic builder) likes to remind me: “Just so you know, none of that stuff you type is private.” So maybe hold off before feeding it every emotional breakdown you’ve had since the Bush administration.

AI Therapy vs Traditional Therapy: What’s Missing?

Still… I’ve seen people get real insights—moments of clarity—from the right prompt. “What are my defenses in this argument?” or “How might my past be influencing this current dynamic?” can lead to moments of reflection that would take longer in a 50-minute session.

However, research indicates that AI therapy chatbots can show increased stigma toward certain mental health conditions and may respond inappropriately to crisis situations like suicidal ideation.

But I’m also deeply saddened by what might be getting lost. Instead of calling a friend to talk through a fight and maybe deepen the relationship, we type into a machine. Instead of messy, vulnerable connection, we seek out tidy digital affirmation. This shift away from human connection and authentic relationships could have long-term implications for our emotional development. Socrates (our beloved OnlyFans philosopher) might be banging his head against the wall over the state of modern discourse.

Covid taught us that we need each other. That we need touch. Real conversation. The kind of presence that doesn’t come from pixels. Just as social media affects our mental health in complex ways, AI therapy tools may be creating new patterns of digital dependency.

According to research published in PMC, while AI shows promise for early identification of mental health risks and treating large volumes of patients, significant concerns exist about bias leading to inaccurate assessments and perpetuation of stereotypes.

The Future of Digital Mental Health

So is ChatGPT a helpful emotional tool? Or another layer of isolation dressed up as insight? Does it give us language for what we feel or quietly flatten it to something easier to digest?

Maybe both. I don’t have a final answer. But I’m deeply curious. Because what we do with our pain and how we make sense of it matters. Whether that’s with a therapist, a friend, a notebook, or yes… a really smart machine.

The key is finding balance. AI therapy tools can be valuable supplements to human connection and professional mental health care, but they shouldn’t replace the irreplaceable elements of human empathy, professional expertise, and genuine relationship. Consider exploring traditional therapy approaches alongside AI tools for a comprehensive approach to mental wellness.

Frequently Asked Questions About AI Therapy

Is ChatGPT a replacement for therapy?

No, ChatGPT and AI tools should complement, not replace, professional therapy. While AI can provide emotional support and self-reflection opportunities, licensed therapists offer specialized training, ethical guidelines, and the human connection essential for deeper healing.

Is it safe to share personal information with AI?

While AI tools can be helpful, remember that your conversations may not be private. Avoid sharing highly sensitive information and consider AI therapy as a supplement to, not a replacement for, professional mental health care.

What are the benefits of AI therapy tools?

AI therapy tools offer 24/7 availability, cost-effectiveness, privacy, and immediate support. They’re particularly useful for journaling, practicing conversations, and learning about mental health concepts. Meta-analyses confirm that computer-aided cognitive behavioral therapy delivered via apps is equivalent to or even more effective than standard CBT for certain conditions.

Can AI therapy help with serious mental health conditions?

For serious mental health conditions like depression, anxiety disorders, or trauma, professional therapy is essential. AI tools can provide additional support but should never replace evidence-based treatment from qualified professionals. Find a licensed professional near you.