What Behavioral Health Providers Should Consider Before Bringing AI Into the Room

By Allison Sikorsky, DNP, PMHNP-BC
Founder & CEO, PMHScribe

When we talk about AI scribes, we tend to discuss what happens after the appointment.

Was the note accurate? How much editing did it require? Did it save time? Did it capture the Mental Status Exam, assessment and plan correctly?

Those are important questions, but in behavioral health, I think we’re missing another one:

What does the presence of AI do to the actual experience of the appointment?

Because an AI scribe isn’t only documentation technology. The moment you introduce it to a patient, it becomes part of the room. And in psychiatry and therapy, the room matters.

The Patient Knows Something Has Changed

Think about what we’re asking patients to do during a mental health appointment: talk about the things they’re afraid of, describe thoughts they may never have said out loud.

Discuss relationships, trauma, substance use, family, anger, grief, medications, intrusive thoughts or the worst week they’ve had in years.

Now imagine beginning that conversation with: “I’m going to use an AI tool during our visit today.”

Even when the technology is secure and the provider is completely comfortable with it, the patient may have questions. What is it listening to? Where does the information go? Is it recording me? What gets put in my chart? Who can see it?

Can I say no?

Those aren’t technology questions. They’re trust questions. And I think behavioral health providers need to treat them that way.

How Should a Mental Health Provider Introduce an AI Scribe to a Patient?

The explanation doesn’t need to become a technology lecture. In fact, it probably shouldn’t.

Patients need a clear, understandable explanation of what the tool does, how it is used during their visit, what happens to their information and what choices they have. Then they need an opportunity to ask questions.

The exact language will vary depending on the technology, organization, workflow and applicable requirements, but the principle is simple:

A patient shouldn’t have to understand artificial intelligence to understand what is happening in their appointment. If we can’t explain the technology simply, we shouldn’t expect patients to automatically feel comfortable with it.

Can AI Actually Make a Mental Health Appointment Feel More Human?

This is where the conversation gets interesting, because technology can create distance, but it can also remove it. Most clinicians know the strange choreography of trying to be fully present with a patient while simultaneously documenting the encounter.

Listen.

Make eye contact.

Ask the next question.

Remember the medication detail.

Type.

Look back at the patient.

Type again.

Try not to make the patient feel like the computer is getting more attention than they are.

We’ve normalized this because documentation is part of the job, but from the patient’s side of the desk, it can feel very different.

When documentation technology works well, the most meaningful change may not be what happens to the note, it may be what happens to the clinician.

More eye contact.
More listening.
Fewer interruptions.
Less attention divided between the patient and the keyboard.

That’s not really a documentation outcome. It’s a patient-experience outcome.

And I think we should start measuring it that way.

How Do Patients Feel About AI Scribes in Mental Health Care?

There probably won’t ever be one universal answer.

Some patients may barely notice. Some may appreciate having more of their clinician’s attention. Some may be curious about the technology. And some may be uncomfortable with the idea of AI being involved in a deeply personal conversation.

All of those reactions are reasonable.

The mistake would be assuming that because the technology feels routine to the provider, it feels routine to the patient. We’ve spent a lot of time talking about clinician adoption of healthcare AI. I think patient adoption is the next conversation.

Not because patients need to become AI experts, but because trust has always been part of the infrastructure of mental health care. If technology changes that relationship, even subtly, we need to pay attention.

What Happens If a Patient Doesn’t Want an AI Scribe?

This is something practices should think through before the situation happens. Not after.

What is the workflow if a patient is uncomfortable? How does the provider respond? Is there an alternative documentation process? Does declining the technology create friction or make the patient feel difficult?

Those details matter.

A patient should never feel like asking a reasonable question about technology is interfering with their care. And a clinician shouldn’t have to invent an answer in real time because the practice never discussed it. AI implementation isn’t just choosing software and training providers to use it.

It’s deciding how the technology will be experienced by the people receiving care.

The Best Technology May Be the Technology Patients Stop Thinking About

I don’t think success looks like patients leaving appointments impressed by the AI. Ideally, they aren’t thinking much about it at all.

They’re thinking:

My provider listened to me.

I didn’t feel rushed.

I didn’t have to repeat myself.

They remembered what mattered.

I felt like there was actually a person sitting across from me.

That is a very different standard for evaluating an AI scribe, and especially in behavioral health, I think it’s the right one. We can measure minutes saved. We can measure documentation completion. We can measure note quality and workflow efficiency. We should.

But we should also ask:

Did the technology give something back to the encounter?

Because the promise of AI in mental health care shouldn’t be that we can fit more technology between clinicians and patients. It should be that we can finally remove some of the technology that has been sitting between them for years.

Frequently Asked Questions About AI Scribes and the Patient ExperienceAre AI scribes appropriate for mental health appointments?

AI scribes can be used in behavioral health settings, but practices should evaluate technology specifically for their clinical environment, privacy and security requirements, documentation needs and patient experience. Mental health conversations often contain highly sensitive information, making thoughtful implementation especially important.

Should patients know when an AI scribe is being used?

Providers and organizations should follow the consent, disclosure and privacy requirements that apply to their specific technology, practice and jurisdiction. Beyond compliance, patients should understand how technology is being used during their care and have an opportunity to ask questions.

Can a patient refuse an AI scribe?

Consent and disclosure requirements can vary by technology, organization and jurisdiction. Practices should establish a clear process for handling patient questions, concerns or preferences before implementing an AI documentation tool.

Can AI scribes improve patient engagement?

One potential benefit of ambient documentation is reducing the amount of attention clinicians must devote to manual note-taking during an encounter. When implemented effectively, that can allow clinicians to spend more of the visit listening and engaging directly with the patient.

What should behavioral health practices consider before introducing an AI scribe?

Practices should evaluate privacy and security, clinical documentation quality, provider workflow, patient communication, consent requirements and what happens when a patient is uncomfortable with the technology. Implementation should consider the entire clinical experience, not only the finished note.

We Built PMHScribe for the Room, Not Just the Record

PMHScribe was created specifically for behavioral health because psychiatric documentation doesn’t happen separately from psychiatric care.

It happens while a clinician is trying to listen. To notice. To think.

To connect.

The technology should respect that.

Yes, the note matters. But so does everything that happens before the note exists.

And if AI can give clinicians even a little more space to be fully present with the person sitting across from them, that may ultimately be one of its most valuable contributions to mental health care.

The future of documentation shouldn’t just be faster. It should feel more human.

---

Learn more about PMHScribe, an AI documentation platform built specifically for psychiatrists, PMHNPs, therapists, counselors and behavioral health organizations.