5 Things Your Patients Want to Know About AI Documentation
Published March 12, 2026
Patients are increasingly aware that artificial intelligence is being used in healthcare. News coverage, social media discussions, and direct experience with AI chatbots have made the technology visible in ways it wasn't even two years ago. When clinicians introduce AI-assisted documentation into their practice, patients have questions — and those questions deserve thoughtful, honest answers.
Based on conversations with clinicians using AI scribes, here are the five questions patients ask most often — and how to address each one in a way that builds trust rather than eroding it.
1. "Is the AI Listening to Everything?"
This is almost always the first question. Patients want to know whether an AI system is passively recording everything that happens in the room — including what they say before the session formally begins, during breaks, or in casual conversation.
The answer should be specific and concrete. With tools like MH Scribe, recording is initiated manually by the clinician — it does not run continuously in the background. The clinician starts the recording when the session begins and stops it when the session ends. There is no ambient listening, no always-on microphone, and no recording outside of the defined session window.
Explaining the manual start-and-stop process reassures patients that they have visibility into when recording is active. Some clinicians make a point of starting the recording visibly, so the patient can see exactly when it begins.
2. "Who Sees My Session Data?"
Patients often worry that their session content might be visible to the software company, other clinicians, insurance companies, or unknown third parties. This concern is especially acute in mental health, where sessions cover deeply personal material.
The straightforward answer: only the treating clinician (and authorized members of their care team, if applicable) can access session data. Under the HIPAA Privacy Rule, covered entities must limit use and disclosure of PHI to the minimum necessary for the intended purpose. At MH Scribe, patient records are scoped to the clinician's account with role-based access controls. The platform team does not review, read, or access individual session transcripts or notes as part of normal operations. Patients also have the right under HIPAA to request access to their own health information, including any AI-generated notes in their record.
3. "Will My Data Train Some AI Model?"
This is the question that has gained the most urgency as public awareness of AI training practices has grown. Patients have read about companies using user data to improve their models, and they want to know if their therapy sessions will end up as training data.
The answer is unequivocal: no. At MH Scribe, patient data is never used to train, fine-tune, or improve AI models. This is not just a policy — it is a contractual guarantee in our BAA and Terms of Service. Session transcripts are processed to generate your clinical documentation and are not fed into any training pipeline, shared with model providers for improvement purposes, or used in any aggregate or anonymized dataset for model development.
This is a point worth stating clearly and directly. Patients should not have to wonder.
4. "Does the AI Make Decisions About My Care?"
Some patients worry that the AI is influencing clinical decisions — choosing diagnoses, recommending medications, or determining treatment plans. This concern reflects broader anxieties about algorithmic decision-making in high-stakes contexts.
The reality is that AI scribes like MH Scribe are documentation tools, not clinical decision-making systems. The AI generates a draft of the session note based on what was discussed. The clinician reviews every word, edits as needed, and makes all clinical decisions independently. The AI may suggest billing codes or flag diagnostic considerations, but these are suggestions for the clinician to evaluate — they are never applied automatically and never communicated to the patient or to payers without the clinician's explicit review and approval.
Framing the AI as a "documentation assistant" rather than a "clinical tool" helps patients understand its limited role in the care process.
5. "Can I Opt Out?"
Patients want to know that using AI documentation is a choice, not a condition of treatment. This matters both ethically and practically — a patient who feels coerced into AI-assisted recording may withhold information that is clinically relevant.
The answer should always be yes. Patients can opt out of AI-assisted recording at any time, for any reason, without it affecting their care in any way. If a patient declines, the clinician simply documents the session the way they did before — by writing notes manually. The therapeutic relationship continues without interruption.
Making opt-out easy and judgment-free is essential. Some patients will want to try it for a few sessions before deciding. Others will decline from the start and may change their mind later. Both are perfectly fine.
How to Bring It Up Naturally
The best time to introduce AI documentation is during onboarding or at the beginning of a session, before recording starts. Keep the explanation brief and matter-of-fact: you're using a tool that helps you take better notes so you can be more present during sessions. Explain what it does, what it doesn't do, and that opting out is always available.
A written consent form can serve as a trust-building tool rather than just a legal formality. When patients see that you've taken the time to document exactly how the technology works, what data is collected, and what their rights are, it signals that you take their privacy seriously. MH Scribe provides a customizable consent form template that covers these points and can be adapted to your practice's needs.
References
- Individuals' Right under HIPAA to Access their Health Information. U.S. Department of Health & Human Services. https://www.hhs.gov/hipaa/for-professionals/privacy/guidance/access/index.html
- Summary of the HIPAA Privacy Rule. U.S. Department of Health & Human Services. https://www.hhs.gov/hipaa/for-professionals/privacy/laws-regulations/index.html
- Business Associate Contracts. U.S. Department of Health & Human Services. https://www.hhs.gov/hipaa/for-professionals/covered-entities/sample-business-associate-agreement-provisions/index.html