AI Psychiatry
Artificial intelligence is revolutionizing how we live, work, and heal. AI is rapidly transforming mental health by expanding access to support, enhancing clinical insights, and personalizing care—yet it also raises urgent questions about empathy, safety, and the irreplaceable value of human connection.
Natural language processing (NLP) is being used in the Sultan Lab to analyze clinical notes and better understand mental health risks, such as e-cigarette use, to create more informed and personalized treatment plans.
SCHEDULE YOUR CONSULTATION TODAY
Reshaping Mental Health
AI is reshaping psychiatric treatment by expanding access through 24/7 support, enhancing personalization with data-driven insights, and improving patient engagement with tools designed to meet people where they are.
At Integrative Psych, we use AI not to replace human care, but to extend it—leveraging technology to make mental health support more immediate, individualized, and effective while preserving the essential role of human connection.
Expanding Access and Availability
- 24/7 support: AI-driven chatbots and digital companions can provide immediate responses during times when human clinicians are unavailable.
- Bridging gaps in care: With millions on waitlists for therapy, AI tools help fill the gap by offering psychoeducation, symptom tracking, and supportive check-ins.
Enhancing Clinical Decision-Making
- Pattern recognition: AI analyzes large-scale electronic health records, speech patterns, or even wearable data to detect early warning signs of relapse, psychosis, or depression.
- Personalized treatment: Machine learning models help tailor interventions to an individual’s behavioral patterns, medical history, and context.
Improving Engagement
- Youth-focused tools: Traditional interventions often fail to engage younger patients. AI-driven platforms—such as gamified digital companions or personalized apps—make therapy more relatable.
- Real-time feedback: Tools like digital pets (e.g., the PAWS project) can provide ongoing motivation and reinforcement outside the therapy room.
Statistics
64%
Greater reduction in
depression symptoms
using AI-based
conversational agents
(chatbots) compared to controls
Hallucination vs. Confabulation:
Rethinking AI Error Terminology
perceiving stimuli that aren’t real
—e.g., hearing voices or seeing things without external input; a fundamentally sensory phenomenon
the production of false or distorted memories
—usually to fill in gaps due to memory impairment—without the intention to deceive
In the AI community, especially when discussing large language models (LLMs), the term "hallucination" has been widely applied to the generation of false or fabricated content—like invented facts or citations.
This usage is misleading: LLMs don’t perceive or hallucinate—they simply produce outputs through probabilistic predictions based on training data. They have no sensory experiences.
A more linguistically precise and conceptually accurate term for these AI errors is confabulation—paralleling how the human brain might fill memory gaps with plausible but false narratives.
This clarification of terminology builds clarity and trust.Calling AI-generated fabrications “confabulations” signals they stem from memory-like gaps, not perceptual anomalies—helping avoid sensationalism and build accurate expectations.
“Confabulation” grounds discussions in cognitive theory.
Scholars in neural networks and computational cognitive science have advocated for “confabulation” as it more accurately reflects “creative gap-filling” processes.
Despite its benefits, AI should remain a supplement
to clinical or therapeutic care.
When thinking about what truly makes something human, we think about an entity having a voice, emotion, a sense of physicality, and much more. There are a number of ways AI companies are incorporating these aspects into their technology: we see it with Tesla's humanoid robot, Optimus, with incorporation of intonation in AI systems, the ability of systems to do sentiment analysis, and the extraction of emotion from plain text.
There is concrete evidence that people are not getting enough connection, with the surgeon general even having declared an epidemic of loneliness. People are lacking connection more than ever, and are turning to AI to fill the void instead of real people. When Chat 4.0 was deleted, there were people on Reddit describing that it felt like a friend of theirs had died. AI companies should be held accountable for some of these outcomes. AI should not be a substitute for human connection.
Users are generally more likely to trust and engage with real people as opposed to technology. A human is more likely to accept advice, follow prompts, or disclose personal data to another human.
AI Changing Patient Care
The impact of AI on the field of psychiatry is going to be much different in nature and scope compared to its impact on other medical fields. In other medical specialties, data is structured and concrete. However, psychiatry and therapy relies on the interpretation of unstructured data on behalf of the psychiatrist, interpretation of: facial expressions, speech patterns, and the patient’s spoken report. AI will help in identifying certain behavioral or spoken word “markers” for certain mental disorders, helping to structure the field of psychiatry.
Another promising development at the intersection of AI and psychiatry is the development of wearable devices that track various patient biometrics, stress level changes, and speech patterns to better inform treatment and help predict when a patient might fall into a psychotic or depressive episode.
We are also seeing an increase in chatbot-based therapy companions. Cannabis Use Disorder is prevalent among young adults, making them more likely to develop certain psychiatric conditions like depression. I’ve found existing digital therapeutics to lack developmental appropriateness and intrinsic motivation for these young adult populations. Generic interventions fail to respond to individual triggers, contexts, and real-time behavioral patterns. In the Sultan Lab, we are working to develop PAWS, a more developmentally appropriate, HIPAA-compliant digital pet companion for data-aware, personalized interventions on cannabis use and depression.
Meet The Founder
Dr. Sultan, son of psychiatrist, is a dual board certified mental health physician who has made it his life’s mission to explore, learn and practice the many treatments for the mind and its troubles.
Dr. Sultan’s path has led him to extensive training in therapies ranging from meditation and mindfulness to practical cognitive behavioral therapy and deep psychodynamic therapy. As a physician, he has explored how classical and innovative medical treatments can foster healing of the mind as well as personal growth.
These extensive ranges of therapies can be effective for many individuals, however they often take years to get results and many individuals hit walls in their treatment—getting only limited benefits.
Dr. Sultan founded Integrative Psych to push past the barriers of traditional mental health treatment by using naturally derived compounds in conjunction with conventional treatments such as medications, lifestyle interventions and psychotherapy to heal the pain and struggles in our minds beyond what traditional mental health treatments can accomplish.
