📙 Download the free episode companion guide https://epg.drbrooklynstormetools.com/

You paste the ChatGPT output into your website, your supervisor loves it, your family loves it, and then someone who looks at private practice websites every week spots the AI patterns and the wrong information hiding between all those long dashes. This episode covers the ChatGPT mistakes therapists make most often and how to use ChatGPT safely in private practice, starting with how to get better ChatGPT output for website copy by giving it a framework, a style, a length and an audience.

I share why you need to verify AI output as a health professional, including an example of a report where the DSM information for a diagnosis was wrong and the author missed it because she was too close to her own work. We also look at why the answers can feel confusing (it mirrors the confusion you type in), why telling it what you do not want tends to backfire, why you can get different answers to the same question, and how unintended plagiarism with AI can happen when the information it gives you was put in by someone else. I also explain how to brainstorm with AI by using it to unpack and explore your own ideas rather than asking it to generate a pile of ideas that other people already had.

Then we turn to your clients, many of whom are already talking to ChatGPT conversationally without knowing how to turn their privacy on and off or that it tends to agree with them and repeat back what they have said. I share why I believe AI psychoeducation for therapy clients belongs inside your duty of care when clients use ChatGPT, how a simple handout or email template with a few guardrails can cover it, and why a note in the file about the handout or conversation is a good idea, especially with clients using ChatGPT for mental health support. I also mention that the codes of ethics I looked up had been updated for 2026, with the social work code the one I could only find in its 2020 edition, so an AI code of ethics for Australian counsellors, psychologists and social workers is something to keep an eye on.

Along the way I give you an update on my task tracker, what a hyperfocus week looks like in the data, and what Etsy's new AI terms meant for my shop.

If you would like fresh eyes on your website, you can book a Website Wellness Check here: https://sales.brooklynstorme.com/wwc-aud/

If you would like to learn how to set up your AI to suit the Australian codes of ethics, you are welcome to join the AI for Therapists community for $7 a month with a 7 day free trial: https://www.skool.com/brooklyn/about

Keywords used: ChatGPT mistakes therapists make, how to use ChatGPT safely in private practice, how to get better ChatGPT output for website copy, verify AI output as a health professional, unintended plagiarism with AI, AI psychoeducation for therapy clients, duty of care when clients use ChatGPT, ChatGPT privacy settings for clients, AI code of ethics for Australian counsellors psychologists and social workers, clients using ChatGPT for mental health support

Mini FAQ:

Q: Should therapists give clients information about using ChatGPT safely? A: Yes, I believe it falls under psychoeducation and your duty of care. A short handout with a few guardrails is enough, and it helps to note in the client file that you gave it to them.

Q: Why do I need to check ChatGPT output before using it in a report? A: Because it can contain wrong information, such as incorrect DSM criteria, and you are responsible for what you share as a health professional. Use it to draft, then verify before you finalise or submit.

Q: How do I get better quality writing from ChatGPT? A: Give it a framework with the style you want, how long or short it should be, who it is for and any specific context. The more specific you are, the better the output.

Q: Why does ChatGPT feel confusing to use? A: It mirrors the input you give it, so confusing input tends to bring confusing output. Tell it what you do want instead of what you do not want, and bring it back to your original question when the conversation drifts.

Q: Why does ChatGPT give different answers to the same question? A: It generates answers from a huge pool of conversations that changes every day, rather than from reliable sources. If you want a factual answer, go to Google instead.