Planning a Mental Health and AI Panel: Speakers, Safety and Evidence
Plan a mental health and AI panel around one audience and one goal. Include qualified clinical input, informed participation by any lived-experience speaker, and written plans for evidence standards, sensitive questions and local crisis resources.


By Harvey Castro, MD, MBA, emergency physician and founder of Signal & Stage.
For a mental health and AI session, define the audience’s main question, select relevant clinical and technical perspectives, and agree on a plan for sensitive audience questions. If someone contributes lived experience, confirm what they want to share and respect their limits.
Speakers in this guide: Amy Kelly, Karin Andrea Stephan, Lupita Almuhana and Taz Hart.
Hypothetical planning example: During a panel about AI and mental health, an audience member describes a family member’s distress. The moderator and event team need an agreed response and a way to connect the person with appropriate support.
Prepare the moderator and event staff for personal disclosures before the session begins.
The following planning suggestions address session scope, speaker selection and audience support.
Ask speakers to distinguish evidence about a specific tool and use case from broad claims about AI. Discuss limitations, privacy and the role of qualified care before presenting a tool as useful.
Which speaker fits a behavioral health session?
Consider Amy Kelly, LMFT when clinicians want to examine where AI fits into daily work and how to protect the therapeutic relationship. Consider Professor Elvira Pérez Vallejos when the committee wants to explore the ethics of AI in therapy, grief and digital wellbeing.
For a 60 minute virtual session, tell the bureau who will attend and the question you want the speaker to address. Confirm the length, availability and fee with the speaker before listing them on the program. Discuss your event with Signal & Stage.
What should the session accomplish?
- Help clinicians decide whether and how to use AI tools in practice Who is in the room: therapists, psychiatrists, primary care, nursing. Consider: panel or workshop.
- Help leaders protect workforce wellbeing as AI changes jobs Who is in the room: executives, HR, managers. Consider: keynote or fireside chat.
- Help patients and the public understand what AI can and cannot do Who is in the room: patients, families, community. Consider: fireside chat in plain language.
Choose one primary goal. If the program covers several audiences, give each a clearly defined session or segment.
Who belongs on a mental health and AI panel?
- A licensed clinician who can say what good care looks like and where AI fits or fails.
- An expert in how the tools are built, sold and adopted.
- A voice of lived experience, with full consent, who decides what to share.
Add a moderator who has done the reading and will stop overclaiming from anyone, including the panel.
On the roster:
- Amy Kelly, LMFT: mental health practice and clinical AI workflows.
- Karin Andrea Stephan: AI and mental health, human trust and adoption.
- Lupita Almuhana: workplace wellbeing and resilience, founder of Stressie.
- Taz Hart: executive communication, leadership presence, storytelling and human connection.
What safety rules does this session need?
- Crisis resources on screen at the start and end. In the US, that includes the 988 Suicide and Crisis Lifeline. Add local resources for international events.
- No detailed descriptions of self-harm methods, in slides, stories or Q&A.
- Patient stories only with consent, and de-identified.
- No outcome claims without evidence. If a speaker says a tool “reduces depression,” ask for the study.
- A plan for Q&A. The moderator should know how to respond kindly and move on, and staff should know where to guide that person afterward.
- Speakers receive these rules in writing.
What questions make the panel useful?
- Where have you seen AI help a patient or clinician in mental health, and what made it work?
- Where have you seen it go wrong?
- What should a clinic ask a vendor before using an AI mental health tool?
- How should a manager talk with a team worried that AI will replace them?
- What should a person in distress know before using a chatbot?
Planner checklist
- The session has one goal.
- The panel includes a licensed clinician.
- Any lived-experience speaker has given full consent and controls their story.
- Crisis resources are on the first and last slide.
- Speakers have the safety rules in writing.
- The moderator has a plan for difficult Q&A moments.
- Every outcome claim has a source.
Frequently asked questions
Is AI ready to be used in mental health care? Evaluate the particular tool, intended use and supporting evidence. The National Institute of Mental Health’s overview of mental health technology identifies effectiveness, privacy and overstated claims as concerns. A panel should explain these limits rather than endorse AI tools as a group.
Does a mental health and AI panel need a psychiatrist? It needs a licensed clinician. That can be a psychiatrist, psychologist, therapist or counselor, depending on your audience.
How should a moderator handle a personal disclosure from the audience? Agree on a response with qualified event support staff beforehand. Acknowledge the person without asking for private details in public, and connect them with the designated support contact. Include an emergency response plan; a panel discussion is not an individual clinical assessment.
Does this topic work for a corporate wellbeing event? Yes. Focus on workforce wellbeing and how AI changes work, with a speaker who works in that setting.
Plan your session with care
Share your audience, the question you want to address and your proposed format. We can discuss relevant speakers and the planning details that still need confirmation.
Related reading: How to choose a healthcare AI speaker | Keynote, fireside chat, panel or workshop? | The human side of AI adoption
Harvey Castro, MD, MBA is the founder of Signal & Stage. The collective helps event planners explore healthcare, AI and leadership speakers.
www.HarveyCastroMD.com | /HarveyCastroMD | #DRGPT
The human side of healthcare AI: topics beyond the panel
A wider conference program can examine therapeutic relationships, digital wellbeing, communication and the experience of patients and caregivers. Choose one of these questions for each session and match it to the speaker’s approved scope.
- Professor Elvira Pérez Vallejos: AI in therapy, responsible AI, grief and digital wellbeing. Read the approved topics and available speaking evidence on the profile.
- Gil Bashe: Healthcare leadership, AI accountability and patient and caregiver experience. Read the approved topics and available speaking evidence on the profile.
Elvira’s approved session AI in Therapy: Ethical Obligations in Practice examines responsibilities around chatbots, confidentiality and clinical scrutiny. Gil’s published topics connect health innovation with accountability, caregivers and system value. Amy Kelly’s clinical AI workflow perspective and Karin Andrea Stephan’s trust and adoption perspective provide other starting points.
A proposed combination of Harvey with Amy or Karin can connect clinical AI with human experience. Confirm participation and final content individually. Use the available recordings on their profiles to distinguish stage excerpts from interviews; a sample does not establish availability.
Explore six conference topic groups
- Practical healthcare AI
- AI governance, risk and trust
- Health innovation and business growth
- Mental health and human experience
- Enterprise AI and organizational change
- Leadership, resilience and innovation
Build a healthcare AI conference agenda or plan a women’s health innovation conference.
Planning routes updated November 3, 2026. Proposed combinations remain subject to each speaker’s confirmed participation, scope, format, duration and language. An inquiry does not reserve a date.
