By The Riverd Team. Reviewed against NCBTMB Standards of Practice and FSMTB Model Practice Act language as of April 2026.
This article is for informational purposes only and does not constitute legal, medical, or insurance advice. Consult your state licensing board, malpractice carrier, and any contracted payer networks for specific requirements.
The short answer: yes, with conditions
AI SOAP notes insurance questions almost always come down to one rule: the licensed practitioner has to be the author and signer. The AI is a drafting assistant, not the practitioner of record. If you review the draft, edit anything that does not match the session, and sign with your name and license number, an AI-assisted note is treated the same as one you typed yourself. The tool that produced the first draft does not change who is responsible for the chart.
The NCBTMB Standards of Practice place documentation responsibility squarely on the licensed therapist. They do not dictate the tool used to produce the words. A handwritten note, a typed note, and an AI-drafted note that you reviewed and signed all clear the same bar: accurate, contemporaneous, and attributable to a licensed clinician. The AMTA's practice resources carry the same framing. Neither body endorses or prohibits AI drafting. They both put the obligation on you.
Take the practical case of a solo LMT in Austin running 25 sessions a week and billing two HSA-pay clients per month. Her notes need to satisfy two audiences: her HSA administrator if a receipt gets questioned, and her own malpractice carrier if a client files a claim. In both cases the question is "did the licensed therapist document this session?" The answer is yes if she read the draft, fixed anything off, and signed with her credentials. You sign it, you own it. That is the entire frame, and the rest of this guide is about making sure the workflow you build actually clears that bar in practice.
What insurance carriers and auditors actually look at
Reviewers care about four things, in this order. First, subjective accuracy: does the client's stated reason for the visit match the rest of the chart and the next visit's intake? Second, objective findings tied to the session: specific muscles worked, palpation findings, range of motion measured, postural observations. Third, treatment plan continuity: does the plan in visit 3 reflect what happened in visits 1 and 2, or does each note read like the first time you ever met this client? Fourth, a verifiable signature with date and license credentials, ideally with an audit trail showing when the note was created and when it was signed.
AI-drafted notes pass all four when the practitioner reads and edits before signing. The AI is good at structuring the SOAP fields and pulling the right clinical vocabulary. It is your job to confirm that the subjective matches what the client said, the objective matches what your hands found, and the assessment and plan reflect the actual conversation about next steps. None of that is hard. It is the work you were going to do anyway. The tool just shaves off the mechanical part of organizing it.
The failure mode is a real one, but it is not unique to AI. Auditors flag charts that look identical session after session: same words, same phrases, same range-of-motion numbers down to the degree. That pattern reads as cloned documentation whether a human typed it or an AI drafted it. The fix is the same in either case. Each note has to reflect the actual session. If you used AI to draft and the output is generic, edit it until it matches. If you typed the note from scratch and copy-pasted from last visit, fix that too. The drafting tool is not the audit point. The chart is.
The practical workflow that keeps you compliant
A clean five-step workflow handles the audit question and saves you time. One, capture the session details right after the client leaves: a 30-second voice memo, three typed bullet points, or a structured intake template. Two, the AI drafts the SOAP structure from your input, organizing what you said into subjective, objective, assessment, and plan. Three, you review and edit, at minimum 30 seconds per note, more if anything in the draft does not match what actually happened. Four, you sign, with electronic signature or initials plus your license number. Five, the note is stored with a timestamp and an audit trail showing who edited what, when.
Steps three and four are the load-bearing parts. The 30-second review is not arbitrary. It is the difference between a chart that holds up and one that reads as auto-generated. If you are spending 10 minutes per note by hand today, an AI draft plus a focused review brings that down to two or three minutes per chart without removing the human judgment that auditors actually look for. The judgment is the point. The typing was never the point.
If your bottleneck is the typing itself, our companion guide on how to write SOAP notes faster breaks down the template, carry-forward, and dictation moves that pair cleanly with the review-and-sign workflow above.
Riverd's AI SOAP Notes draft a structured note from a 30-second voice memo. You review, edit, and sign. Free up to 20 appointments a month. See how it works.
This is one example of a tool that runs the workflow above end to end. There are others. The shape of the workflow matters more than the brand: voice or text input, AI draft, practitioner review, signed note, timestamped storage. If you are evaluating tools, ask vendors how they handle the signature and the audit trail. Those two pieces are what a reviewer will actually inspect, and they are the two places where a tool either supports compliant practice or quietly makes it harder.
The two cases where AI notes can get you in trouble
Be honest about the failure modes. Case one: high-volume cash practices where the therapist signs without reading. If you are running 30 sessions a week and clicking through 30 AI drafts in a minute, the notes will look identical because you never edited them. That is a documentation problem, not an AI problem, but AI makes it faster to fail. The fix is the review step. Build it into your workflow as non-negotiable, and if a draft is wrong, edit it before you sign. If you do not have 30 seconds per note to read and edit, you are running too hot regardless of the tool, and the chart will tell that story to the next reviewer who looks at it.
Case two: jurisdictions or specific payers that require attestation language. Some auto-injury claim networks and a handful of state boards now ask for a line stating the note was "personally documented by the treating provider." That language exists because of exactly the concern above. Regulators worry about fully delegated documentation. If you bill into one of those networks, add the attestation to your note template before your first claim. The Federation of State Massage Therapy Boards (FSMTB) maintains Model Practice Act language on records that your state board may have adopted in part. Check your state's actual rules, not just the model.
International readers: the framing in this post is US-specific. UK, EU, and Canadian practitioners face different documentation rules and data-protection regimes, and the answer there can shift on consent language, GDPR storage rules, and provincial requirements. Use this article as a starting point, then check your own regulator and any professional association you belong to. In every jurisdiction, the question to keep asking is the same: did the licensed therapist read this note and stand behind it? If you can answer yes, an AI draft does not weaken the chart. Never assume the tool guarantees claim acceptance. The variable that gets a claim paid is the documentation itself.
Key Takeaways
- AI SOAP notes are acceptable for insurance when the licensed practitioner reviews, edits, and signs every note.
- Auditors care about authorship, accuracy, and continuity. The drafting tool is not the audit point.
- Build a five-step workflow that puts a real review step before the signature, and keep it to at least 30 seconds per note.
- Some payers and state boards require attestation language. Check your state board guidance and any contracted networks before you bill.