A flagship moment on this theme
"When someone expresses distress, mentions self harm or is in crisis, it is the same idea as a crash test for a car. You do not wait to find out if it is safe by putting real people in danger. You test it first under the hardest conditions you can create so you can know exactly where it fails."
The crash test analogy is the spine of the episode and Rachel states it crisply before the interview begins. It is quotable on its own and frames the entire conversation for a reader who has not heard the episode.
Where this is a central topic
9 episodes
Crash-Testing AI for Mental Health with Shirali and Arul Nigam
Three research articles frame the conversation: AI chatbots systematically violating ethics standards, users developing delusions after chatbot interactions, and teenagers dying by suicide after AI companion engagement.
Tiered Care, Technology, and the Future of Mental Health
The episode maps specific technology applications across five distinct stages of care drawn from a PMC peer-reviewed article, giving the discussion a research-grounded structure rather than a speculative one.
Big Mental Health Stories & Trends to Watch in 2026
Rachel discusses at length the documented harm from AI-only therapy models, active lawsuits, and the case for human-supervised AI handling scheduling, notes, and between-session support.
The Future of SMI Treatment with Dr. Scott Feers
AI tools including point-of-care assistants, LLM-based treatment goal tracking, relapse detection algorithms, and AI scribes are discussed in concrete clinical terms with specific examples.
AI and Measurement-Based Care with Dr. Dylan Ross
Blueprint's AI scribe and AI-assisted clinical decision support are described in detail, with the finding that AI scribing drove a 60 percent increase in MBC adoption named as the episode's sharpest and most surprising data point.
Navigating Tech in Private Practice with Uriah Guilford
The episode's central thread: Uriah's predictions about AI avatars, AI receptionists, and AI advisory teams; Rachel's clinical argument against AI scribes; and the question of where AI helps versus harms in the clinical relationship.
Preparing for the AI Shift in Therapy with Dr. Jordan Harris
The episode's central argument throughout: AI is already outperforming some therapists at validation and empathy-signaling, clinicians have a narrow window to upskill, and the right response is engagement rather than resistance.
When AI Meets Therapy: Risks, Ethics, and Advocacy with Dr. Ajita Robinson
The episode opens with documented harms including a teen suicide linked to AI therapy use, ChatGPT psychosis, and the absence of crisis protocols in most AI chatbots, then returns to this thread throughout.
Real Change
The group returned repeatedly to chatbot therapy as a specific harm — no licensure, no human experience, data collected to train the model — framing it as the sharpest edge of a broader problem with unregulated tech in clinical care.
Where this comes up substantially
2 episodes
Technology, Therapy, and the Future of Care
Rachel argues that the most defensible use of AI in clinical settings is reducing documentation, scheduling, and billing load so clinicians can spend more time doing the actual therapeutic work they were trained for.
Using Brain SPECT Imaging for Mental Health Treatment | Dr. Rishi Sood of the Amen Clinics
Dr. Sood envisions AI coaches filling the gap between appointments; Rachel pushes back on whether that reduces human connection; they find common ground around AI as cognitive-load reducer rather than clinician replacement.
Mentions
2 episodes
Mentions
2 episodes
Understanding Mental Health Parity with Deborah Steinberg
In a post-recording conversation, Deborah and Rachel discuss AI being used to deny insurance claims and the absence of real privacy protections for mental health apps that fall outside HIPAA jurisdiction.
Scaling Mental Health Organizations with Ryan Dewey Smith
Brief mention of Palantir Foundry for data warehousing, KPI tracking, and occupancy trends, framed as a back-office efficiency tool rather than anything clinical.
Looking to go deeper in your own work?
These TSTI trainings build on conversations from the episodes above.