AI Will Make Mental Healthcare More Human


The perception of Artificial Intelligence (AI) in mental healthcare is often that it exists in a background role, like in scheduling appointments or billing, with no direct interaction with patients. However, as technologies progress, more opportunities exist to leverage AI into something more than just a “back-of-house” tool with supporting functions. This article will focus on this “front-of-house” AI.

Many people, possibly influenced by sci-fi movies or watching automated robots on social media platforms, are wary of patient-facing AI in healthcare. They believe it will displace human interaction or undermine employment opportunities for trained mental health professionals. This belief couldn’t be further from the truth.

AI, in the right hands, can actually enhance humanity within healthcare. By interacting directly with patients, reaching underserved communities, and performing much of the heavy lifting for mental health providers, AI allows clinicians to focus more time on the fundamentally human aspects of care delivery.

A human touch requires time; AI creates it

After an initial clinical assessment session (lasting anywhere from 1 to 2 hours), a typical psychotherapy session lasts around one hour, with a single therapist able to see seven to eight patients per day. Assuming the patient shows up (with studies showing the rate of missed appointments at over 40%) [1], they spend time in session sharing an update on what has happened since their last visit, talking through their experiences, thoughts, and feelings. Through structured conversation, the clinician’s job is to continuously assess the patient’s issues, relate this to the clinical literature and contextual knowledge, update the treatment plan if necessary, and ensure the patient is well-equipped to advance through the treatment steps, identifying potential barriers.

That’s a lot to get done in one hour! What if a well-trained AI agent could take on some of the burden? Large language models like ChatGPT have shown us that AI can hold highly naturalistic structured conversations with a human user. It stands to reason that clinically validated AI could work with the patient between sessions, gathering critical information, monitoring progress, ensuring treatment adherence and session attendance, and monitoring for signs of clinical risk or deterioration. By the time the patient next sees their therapist, the therapist can have already received a succinct clinical note with a comprehensive summary of the patient’s progress since the last session.

Rather than draining humanity from therapy, AI will flood the system with more time. A one-hour session could be spent focusing on building a real human connection between the therapist and the patient—empathy, active listening, relationship-building, trust, and expectation management. This blended approach to care would retain the clinical expertise of qualified therapists, relaying information more effectively and efficiently between provider and patient.

Humanity for the many, not the few

While the skeptic would be forgiven for thinking patients might refuse to work with an AI agent, the research tells a different story. For example, a recent study showed that patients were more likely to refer themselves to care if they were greeted by an AI [2], citing personalization, empathy, and customized clinical interfacing as greatly impacting the patient’s positive experience.

AI holds the potential to improve the user’s experience and onboarding into health systems while reducing the stigma of mental health by offering a fundamentally nonjudgmental point of entry. Fascinating was the fact that minority groups benefited most from this technology, demonstrating the potential of digital tools to improve access for underserved communities.

Making mental healthcare a more humane place to work

Unfortunately, burnout is a significant problem for psychotherapists, which became especially evident during the COVID-19 pandemic. According to the American Psychological Association, almost 50% of psychologists believe that they cannot meet the treatment demand from patients and have an overwhelming feeling of being burned out [3].

Psychotherapists have become so overburdened with such poor work-life balance that many have decided to leave the industry. While most are still passionate about helping people, these providers cannot continue due to the incredible stressors they are facing.

To curb this challenge and increase industry retention, mental health providers can use patient-facing AI to do much of the heavy lifting, freeing them up to tackle only those tasks that require a true human expert. This will allow therapists to take on more of a clinical supervisor role and operate at the very top of their capabilities without feeling overwhelmed.

Humanity evolves through technology

It’s indisputable that humans connect intricately through technology, from mobile phones and tablets to laptops. Society has evolved in tandem with technology. Just as the wider healthcare system benefitted from the inventions of medical imaging, genomic data, or even 3D printing prosthetics, implants, and replacement joints, walling off mental healthcare to AI and machine learning would isolate our industry and ignore a foundational piece of the human experience of the very people we aim to help.

AI technology in the mental healthcare sector is representative of the journey from steam to diesel engines in the locomotive sector (or, better yet, the advancement from diesel to battery-powered vehicles) [4]. AI promises to usher in a desperately needed paradigm shift to an industry that still operates solely on overstretched human labor.

We must not allow providers or patients to be left behind. We must embrace these new approaches and guide them towards the betterment of humanity.


Leave a Reply

Your email address will not be published. Required fields are marked *