For practitioners

Can practitioners ethically use AI to write their content?

Yes, for content. Not for clinical reasoning, not for anything touching a patient's record, and not for anything published without you reading it first.

I'm an acupuncturist and functional medicine practitioner. Twenty-seven years of practice, and I now build AI systems for my own work. So this isn't a technology company telling you it's fine. It's a colleague telling you where the line is.

Where is the actual line?

The question people ask is whether AI is ethical. The more useful question is whose judgment is being replaced.

Writing an Instagram post about histamine intolerance isn't clinical judgment. It's education you've delivered a thousand times in a treatment room. Having a tool draft it from your own words, then editing it before it goes out, replaces typing. It doesn't replace thinking.

Deciding what a patient's labs mean is clinical judgment. That doesn't go to a tool. Not because it's forbidden, but because you're the one licensed and accountable for it.

The line isn't really about AI at all. It's the same line you already apply to a VA who drafts your newsletter, or an intern who writes your handout.

What about accuracy? These things make things up.

They do. Confidently, without flagging it, and that's the real risk for our field specifically. A fabricated mechanism or a dosage that reads plausibly is far worse than an obviously wrong sentence.

The structural fix: never let it generate clinical content from its own knowledge. It should write from your material. Your notes, your protocols, your past posts, your own explanations. And everything gets read by you before it exists publicly. Set up that way, the failure mode mostly disappears, because it isn't sourcing facts. It's rearranging yours.

If you're using a general chatbot to write clinical content off a one-line prompt, you will eventually publish something wrong under your license. That isn't hypothetical.

Do I have to disclose that I use AI?

No rule requires it for marketing content, and I don't think you owe anyone a disclosure that a draft was typed by software before you rewrote it. You don't disclose your word processor.

What you owe is that the content is genuinely yours: your reasoning, your voice, your clinical position, reviewed by you. If what's going out isn't something you'd defend in front of a colleague, disclosure isn't the problem. The content is.

What about my patients' information?

Never. No patient names, no case details, no intake forms, no labs, nothing from a chart. Not into a chatbot, not into a content tool, not "de-identified" into a general model. That's a HIPAA conversation with real consequences, and it's the one place I'd tell a colleague to be rigid rather than thoughtful.

If you want to write about a case, do what you've always done. Compose a fictional composite that illustrates the pattern, and write it yourself, from your own head. More on the HIPAA specifics here.

What this looks like in practice

I run a system that drafts my content from my own clinical writing, in my voice, and holds everything for my approval before anything publishes. I built it for my practice first.

This page is Brie Wieselman's own position on where the line sits for a licensed practitioner using AI in marketing, not legal or regulatory guidance. It reflects how she runs her own practice. Reviewed August 2026.

Two ways to get this

Done-for-you: I build and run the system on your practice. $3,000 build plus $400 a month at the founding rate.

The cohort: six weeks, small group, you build your own with me guiding. $1,500 founding rate.

Not ready for either? The free guide walks through the whole method.