Two tools, one label
"AI for behavioral health" now covers a wide range of products. Two of the most common are AI intake and AI documentation. They are often discussed as if they were the same thing. They are not. They solve different problems, at different points in the client journey, and they handle different kinds of information.
Getting the distinction right matters because it changes the questions you ask a vendor, the safeguards you need, and the order in which a practice should adopt them.
What AI intake does
See what this looks like for your practice
In 30 minutes we will identify where your practice is losing clients, quantify the revenue impact, and give you the single highest-leverage fix to implement first.
Book a Strategy Call →AI intake works in the space between "a person reaches out" and "a clinician sees that person in session." It handles logistics:
- Answering calls, forms, and messages, including after hours
- Collecting contact details and the reason for reaching out
- Checking insurance eligibility
- Offering and confirming appointment times
- Routing inquiries by specialty, location, or insurance type
- Following up with people who did not book
- Routing crisis language immediately to human staff and crisis resources
None of that requires clinical judgment. An intake assistant does what a well-trained front desk coordinator does, at hours and volumes a coordinator cannot cover alone. For more on where that line sits, read why AI intake is not the same as AI therapy.
What AI documentation does
AI documentation tools, often called AI scribes or ambient documentation, help clinicians produce clinical documentation such as progress notes, typically from a session recording or the clinician's input. The clinician reviews and finalizes the note.
These tools work with clinical content: what was said in session, symptoms, interventions, and clinical impressions. That places them much closer to clinical care than intake tools, even when the clinician remains responsible for the final record.
Side by side
| AI intake | AI documentation | |
|---|---|---|
| Problem it solves | Lost inquiries, slow responses, insurance uncertainty | Time clinicians spend writing notes |
| When it works | Before the first session | During and after sessions |
| Who uses it | Prospective clients and front desk staff | Clinicians |
| Information it handles | Contact details, reason for reaching out, insurance, scheduling | Session content and clinical documentation |
| Clinical judgment involved | None; anything clinical is routed to a person | Clinician reviews and owns every note |
| Business Associate Agreement | Required | Required |
Different data means different safeguards
Both categories handle protected health information, so both require a signed Business Associate Agreement before any client data is shared. Beyond that, the questions differ.
For intake AI, the key questions are about speed, routing, and boundaries: does it answer every channel, does it check eligibility for your plans, and does it route crisis language to a person immediately? Our HIPAA checklist for AI intake vendors covers the compliance side.
For documentation AI, the content is more sensitive, so the questions go deeper:
- Is session audio stored, and if so, where and for how long?
- Is any session content used to train AI models?
- How does the clinician review and correct a draft before it becomes part of the record?
- How do you obtain and document client consent to recording?
- How does the tool keep psychotherapy notes separate, given the extra protection HIPAA gives them?
Our guide to HIPAA-compliant AI for behavioral health explains why no AI tool is compliant on its own, and what compliance actually depends on.
Which should a practice start with?
Both can be valuable. The question is which problem is costing you more right now.
For most practices, the larger and more measurable losses happen before a client ever reaches a clinician.
Most group practices that measure their intake sit 15 to 25 points below high-performing benchmarks, and for a 200-client practice, closing that gap is worth an estimated $40,000 to $120,000 a year. Those are clients who never arrive, so no amount of documentation efficiency recovers them.
Intake AI also carries a lower clinical risk profile, because it never touches session content. That makes it a natural first step for practices that are new to AI, and it builds the operational foundation that later tools depend on. See the five levels of AI readiness to assess where your practice sits.
BHSAI focuses on administrative AI: intake, eligibility checks, scheduling, and routing. For the full picture of where AI fits in a practice, read Behavioral health AI: a guide for practice owners.