Download the Report
Enter your details below to get instant access to the full report.
A few things clinicians told us
30%
of clinicians now use AI for at least one part of their work, most often documentation.
33%
say they would use AI more if they trusted how patient data is handled.
78%
believe AI should stay out of the therapeutic relationship itself, and draw a firm line at clinical judgment.
Seven chapters, built to be read and acted on.
01
The state of adoption in 2026
Where practice actually is with AI, by setting, role, and task.
02
What clinicians trust, and what they don’t
The tasks professionals will hand over, and the ones they will not.
03
Patients on AI in their care
What people in treatment say, and what they expect to be told.
Privacy, consent, and the data question
The concerns that shape adoption, and what responsible handling looks like.
05
Keeping the human in the room
Adopting AI without eroding the relationship at the heart of care.
06
A practical framework for responsible adoption
A step-by-step guide clinicians and owners can apply this week.
07
What comes next
Where the field is heading, and how to prepare without overreacting.
Written for the people doing the work.
Therapists and counselors
Understand how peers are using AI, and where the profession is drawing its lines.
Psychiatrists and prescribers
See where AI is genuinely useful in a clinical workflow, and where caution is warranted.
Practice owners and leaders
Get a framework for adopting AI across a team responsibly, with privacy and consent handled well.
— The Berries clinical team
28%
Confident their tools protect privacy
64%
Want stronger data safeguards
Responsible-AI frameworks remain the top requested resource.
45%
Save 5+ hours every week
12%
Save more than 10 hours
How we gathered this
This report is based on responses from [N] mental health clinicians surveyed in [month/year], across solo practice, group practice, and clinics in [region]. It also draws on [patient survey / anonymized product data / interviews], and was reviewed by [practicing clinicians / named advisors] before publication.
All data is reported in aggregate. No individual clinician or patient is identifiable.
Get the report