How AI Tools for Therapists Are Built: Trust, Data Privacy, and Choosing What to Adopt - An Interview with Ian Knox and Megan Toomey of SimplePractice

How AI Tools for Therapists Are Built: Trust, Data Privacy, and Choosing What to Adopt - An Interview with Ian Knox and Megan Toomey

Ian Knox and Megan Toomey of SimplePractice take therapists behind the scenes of how AI tools for mental health are actually built, trained, and kept secure.

We're in the middle of AI month, and Curt and Katie wanted to move past the buzzwords and talk with the people who actually engineer this technology. As part of the show's partnership with SimplePractice, they sit down with Ian Knox, Chief Product Officer, and Megan Toomey, Sr. Director of Clinical Support AI Product Management, to pull back the curtain on how AI tools for therapists get designed, what "training the model" really means, and how client data is handled.

They get specific about what to evaluate before adopting any AI tool, from vendor trust and HIPAA compliance to data practices, and why note taking is the most mature use case while insurance, scheduling, intake, and referral matching are still emerging. Ian and Megan also take on the fear that AI-first companies want to replace therapists, and explain why clinicians have to stay at the center of care and review anything they put their name on.

This is a grounded, practical conversation for any therapist trying to decide what AI belongs in their practice, and what to be cautious about, without panic or hype.

In this episode, we discuss:

- How an EHR decides which clinician tasks AI is mature enough to help with

- What to evaluate before trusting an AI vendor with client data

- Why "HIPAA compliant" is a floor, not proof of strong security

- What "training the model" does and does not mean, and why SimplePractice says it is not training an LLM on your data

- How transcripts are retained, and the new opt-in for de-identified data

- Why you remain responsible for every AI-assisted note you sign