Guardrails for counselors and therapists

When you finish, you can explain what never goes into an AI tool and how to protect clients.

Confidentiality first

Client names, stories, session content, notes, recordings, photos, diagnoses and even the fact that someone is a client are confidential. Do not put them into a consumer AI tool unless your employer or practice has approved that tool for that data and any required consent is in place. Privacy laws such as HIPAA in the US and similar laws in many countries apply, along with your ethics code.

Removing a name is not always enough. A unique story can identify a person to someone who knows them. If you would not say it in a public room, do not paste it into a general tool.

Your ethics code and your board

Your profession’s ethics code and licensing board set rules on confidentiality, competence, record keeping, advertising and technology. Read them again with AI in mind. Ask your supervisor or association about current guidance.

Informed consent matters. If you plan to use any AI tool that touches client information, clients should know in plain language and be free to say no.

Accuracy, harm and bias

AI can produce advice that sounds caring and is wrong or harmful. It can miss cultural context, assume a gender or family type, and use language that stigmatizes. Never hand an AI-drafted handout to a client without reading every word.

Chatbots are not crisis tools. If a client asks about using AI for support, discuss it openly and make sure they know how to reach real help in an emergency.

Disclosure and saying no

If AI helped with a handout, a policy or a website, be ready to say so if asked. A line like “I drafted this with AI help and revised it” keeps things honest. For client-facing tools, follow informed consent rules.

Say no to requests to feed real cases into a chatbot, to use AI for diagnosis, or to skip your review. A calm line works: “I can use AI for templates, but not for client information.”

A worked example: the notes shortcut

A therapist is behind on notes and considers pasting session details into a chatbot to polish them. Weak move: pasting the story. Better: write notes in your approved record system, using your own structure. If you want wording help, ask for generic sentence frames like “Client reported [topic] and we discussed [skill],” then fill them in yourself.

You kept the content in the secure system and still saved time on phrasing. That is the pattern: use AI for the shape, keep the substance protected.

Practice now

Reread one section of your ethics code on confidentiality and write how it applies to AI in your own words.
Create a no-identifier sentence frame for a note, and note what you would add inside your approved system.

Use any AI assistant you have. Do not paste private, client or patient information into it.

Check yourself

1. Which is a confidentiality risk?

2. A client uses a chatbot for emotional support. What is a helpful response?

3. What should you do before using an AI tool that touches client data?

Ask the AI tutor

Ask anything about this lesson, in any language. The tutor is an AI: it can be wrong, and it never needs your personal details.

Next lesson: Grow with AI: your 30-day plan

AI-drafted. Checked by an automated validator and a separate AI reviewer; a person has not read every lesson. Translations and tutor answers are AI-made. Found a mistake? Tell us. AI can be wrong. General information only, not professional advice. Not affiliated with any AI company named. Current University is not accredited.