AI for Mental Health: What Therapists Really Think About Digital Support

AI for mental health is becoming increasingly relevant in a world where nearly 50 percent of individuals who could benefit from therapeutic services are unable to reach them. This accessibility gap, combined with the fact that more than half of US psychologists didn’t have openings for new patients in 2024, highlights why digital solutions are gaining attention.

When we look at current usage patterns, the picture is mixed. About 43% of mental health professionals use AI primarily for research and report writing, while 28% of clients use these tools for quick support or as a personal therapist. Despite concerns, the results can be promising—more than 85% of patients reported finding AI therapy sessions beneficial, with 90% expressing interest in using virtual therapists again.

However, the technology isn’t without risks. AI therapists have shown increased stigma toward conditions like alcohol dependence and schizophrenia compared to depression. Furthermore, although most users find AI generally beneficial, specific harms and concerns were experienced by nearly half of both clients and mental health professionals.

How AI is being used in mental health today

The practical applications of AI for mental health are expanding rapidly, addressing critical gaps in care delivery. Currently, AI technologies are being integrated into various aspects of mental health services, from direct patient interaction to behind-the-scenes clinical support.

Chatbots and virtual therapists

AI-powered virtual therapists are showing promising results in clinical trials. In a groundbreaking study from Dartmouth College, participants using an AI therapy bot reported significant symptom improvements—a 51% reduction in depression symptoms and a 31% decrease in anxiety. Notably, users developed strong relationships with these digital therapists, accessing them frequently, even during middle-of-the-night insomnia episodes. Similarly, a Cedars-Sinai study found that over 85% of patients with alcohol-associated cirrhosis considered their AI therapy sessions beneficial, with 90% expressing interest in future virtual therapy.

Mood tracking and journaling apps

AI-enhanced journaling tools are transforming how people monitor their emotional wellbeing. Applications like Grow Therapy have introduced AI-powered journaling that identifies key themes from patient entries, which can be shared with providers before appointments. Another tool, Resonance, offers action-oriented suggestions based on past memories, significantly reducing depression scores in a two-week study. These applications analyze mood fluctuations, offering insights into potential triggers and personalized recommendations for coping strategies.

Administrative support for clinicians

For mental health professionals facing burnout, AI offers substantial relief. Documentation automation technology can generate up to 80% of progress note content, reducing provider documentation time by more than 70%. Additionally, AI tools handle scheduling, appointment reminders, and insurance claims processing. This automation allows therapists to focus more on meaningful client interactions rather than paperwork, potentially alleviating one of the biggest causes of provider stress.

AI-assisted diagnosis and treatment planning

AI excels at processing extensive clinical data to support diagnostic decisions and treatment planning. These systems can analyze electronic health records to identify individuals at risk for developing mental health conditions, enabling earlier intervention. Through natural language processing, AI can identify and classify suicide ideation in patient records. Furthermore, AI provides clinicians with real-time insights about symptom patterns and suggests personalized treatment approaches based on evidence-based indicators.

What therapists really think about AI support

Therapists across the profession are grappling with the rapid emergence of AI in mental health care, expressing varied perspectives on its proper role and limitations. Their viewpoints offer critical insights into how these technologies might best serve patients without compromising care quality.

Mixed feelings about AI replacing human connection

Professional clinicians express significant reservations about AI-based therapy, with a majority (56%) indicating they would never consider accessing therapy from a machine therapist, while 35% remain uncertain, and merely 9% express openness to the idea. Therapists particularly worry about the bond aspect of the therapeutic alliance, which represents the personal connection between therapist and client. As one psychotherapist explained, “AI cannot replicate genuine human empathy and there is a risk that it creates an illusion of connection rather than meaningful interaction”.

Support for AI in non-clinical tasks

Conversely, mental health professionals show enthusiasm for AI handling administrative responsibilities. Many therapists appreciate how AI can streamline logistics tasks like billing client insurance or serving as a “standardized patient” for training purposes. One psychologist aptly summarized this perspective: “I leverage AI to handle the predictable paperwork so I can be fully present and creative in my actual therapeutic work”. Indeed, automation of documentation and scheduling allows practitioners to focus more energy on direct client care.

Concerns about overreliance on AI tools

Nonetheless, clinicians express caution about potential risks. Data security stands as a primary concern, with one therapist noting that “while disclosures and consents can outline how user data will be secured, we have witnessed numerous data breaches among previously trusted tech companies”. Moreover, mental health professionals worry about AI’s inability to recognize emotional nuances in expression or hold space for grief. The American Psychological Association has specifically raised alarms about unregulated AI therapy bots posing dangers to vulnerable individuals.

Therapist views on AI therapy effectiveness

Regarding effectiveness, the consensus leans toward skepticism. Most therapists disagree that AI can effectively reduce mental health stigma or provide genuinely effective support. Many believe  we should limit AI  to training contexts rather than actual therapy. Nevertheless, some acknowledge AI’s potential usefulness in specific domains, such as supporting journaling, reflection, or coaching in less safety-critical scenarios.

Risks and limitations of AI in mental health care

Despite promising advancements, research reveals serious shortcomings in AI mental health applications that warrant careful consideration.

Bias in AI responses and training data

Studies show AI chatbots demonstrate increased stigma toward conditions like alcohol dependence and schizophrenia compared to depression. This bias appears consistent across different AI models. At Cedars-Sinai, researchers found alarming patterns of racial bias in treatment recommendations, with AI platforms suggesting different psychiatric treatments when African American identity was stated or implied. In fact, these biases were most evident in cases of schizophrenia and anxiety.

Inaccurate or harmful advice in crisis situations

In critical situations, AI can dangerously miss the mark. Stanford University researchers found that in 20% of cases involving delusions, hallucinations, or suicidal thoughts, AI failed to provide clinically appropriate responses, whereas licensed therapists provided appropriate responses 93% of the time.

Lack of empathy and emotional nuance

AI fundamentally lacks emotional intelligence and nuanced understanding required for therapeutic connections. Unlike human therapists, chatbots cannot hold space for grief, discern nonverbal cues, or co-regulate with distressed clients. A 2023 study found users often disengaged from mental health chatbots after a few sessions due to the absence of meaningful emotional feedback.

Privacy and data security concerns

Once you shared health information  with AI systems, the tool can potentially disclose it to both intended and unintended audiences. Many mental health applications operate in a “regulatory gray area” outside HIPAA protections. Consequently, sensitive disclosures about trauma, suicidal thoughts, or substance use can be stored, analyzed, and even used for commercial purposes without proper informed consent.

Ethical concerns around AI counseling

Unlike licensed therapists bound by professional codes, there’s no accountability mechanism when AI gets something wrong—no license to revoke, no malpractice recourse. OpenAI itself acknowledged that AI chatbot behavior “can raise safety concerns—including around issues like mental health, emotional over-reliance, or risky behavior”. Without proper oversight, vulnerable individuals face potentially devastating consequences.

Conclusion

AI mental health tools stand at a crossroads between promising innovation and significant limitations. Nevertheless, the human element remains irreplaceable in therapeutic relationships. Most practitioners view AI as a helpful co-pilot rather than a suitable replacement for genuine human connection.

Ultimately, mental health care requires both technological innovation and human wisdom. While AI can process vast amounts of information quickly, only human therapists can truly understand the complexity of human suffering. The question isn’t whether AI will transform mental health care—it already has—but how we ensure these powerful tools serve rather than undermine the deeply human process of healing. If you want to learn more join our Master degree.

FAQs

Q1. How effective is AI in providing mental health support? While AI tools show promise in certain areas, their effectiveness varies. Some studies report positive outcomes, with over 85% of patients finding AI therapy sessions beneficial. However, AI lacks the emotional intelligence and nuanced understanding that human therapists provide, making it more suitable as a complementary tool rather than a replacement for traditional therapy.

Q2. What are the main concerns therapists have about AI in mental health care? Therapists express concerns about AI’s inability to replicate genuine human empathy, potential overreliance on AI tools, and risks associated with data security and privacy. Many worry that AI might create an illusion of connection rather than meaningful therapeutic interaction, especially in handling complex emotional situations.

Q3. Can AI completely replace human therapists? No, AI is not expected to replace human therapists entirely. Most mental health professionals view AI as a helpful assistant for administrative tasks and research, but not as a substitute for the human connection and empathy crucial in therapy. The therapeutic alliance built on shared humanity remains irreplaceable by technology.

Q4. What are some practical applications of AI in mental health today? AI is currently being used in various ways, including chatbots and virtual therapists for immediate support, mood tracking and journaling apps, administrative support for clinicians (like documentation automation), and AI-assisted diagnosis and treatment planning. These tools aim to enhance accessibility and efficiency in mental health care.

Q5. How is the future of AI in therapy shaping up? The future of AI in therapy is likely to be a hybrid model where AI serves as a co-pilot to human therapists. It can help to enhance personalization, improve accessibility, and assist in training and simulations for mental health professionals. However, there’s a growing emphasis on the need for regulation and therapist oversight to ensure responsible implementation and protect patient safety.