Should Your Therapy Website Mention AI? A 2026 Disclosure Guide for Therapists

A year ago, most therapists treated AI on their website the way they treated their EHR vendor: a quiet operational fact, not something to advertise.

In 2026, that's no longer a defensible position.

The APA published formal Ethical Guidance for AI in the Professional Practice of Health Service Psychology in June 2025 and a public health advisory in November 2025. The ACA released parallel recommendations for counselors. The APA also asked the FTC to investigate AI chatbots posing as therapists after well-documented safety incidents involving minors. State licensing boards are following.

Translation: If you use AI in your practice — even just for notes — disclosure is no longer a marketing question. It's an ethics requirement.

The marketing question is what comes next: how you mention AI on your website now shapes whether prospects trust you, whether you stay on the right side of your licensing board, and whether you accidentally signal something you didn't mean to.

This guide gives you the framework.

The 2026 Verdict

• If you use clinical or clinical-adjacent AI (notes, transcription, treatment-planning aids, intake summaries): you must disclose it. The website is one of the right places to do that.

• If you use administrative AI only (scheduling, marketing copy, billing): disclosure is optional but increasingly expected by privacy-aware clients.

• If you embed an AI chatbot that talks to prospects about mental health: stop. This is the highest-risk decision a therapy website can make in 2026.

• Tone matters. "AI-powered therapy" reads as a red flag. "Human-led care, with AI used only for [specific admin tasks]" reads as careful and competent.

• The honest framing — human relationship at the center, AI only where it makes you safer and more available — converts better and protects you better than either silence or hype.

• 
  

The 2026 Landscape: Why This Conversation Changed

Three forces made AI disclosure a 2026 issue rather than a 2027 issue:

1. The APA and ACA both published formal guidance. The APA's June 2025 ethics document is explicit: AI should augment, not replace, human decision-making; psychologists remain responsible for final decisions; and use of AI must be transparent and consistent with HIPAA and state confidentiality requirements. The ACA's parallel recommendations for counselors echo the same principles — transparency, client welfare, professional accountability.

2. The APA issued a public health advisory in November 2025. The advisory specifically addressed generative AI chatbots and wellness apps in mental health, urging clinicians to proactively ask patients about their use of these tools. It also pushed for systemic safeguards and clinician training on bias, privacy, and responsible AI use.

3. Public trust took real hits. A Brown University study published in early 2026 found that AI chatbots, even when prompted to act as therapists, "routinely break core ethical standards of mental health care" and handle crises poorly. The APA's own February 2025 request to the FTC to investigate AI chatbots posing as therapists followed at least one widely reported tragedy involving a teenager and an AI companion. These stories are now in the cultural water — your prospects have read them.

The combined effect: clients are simultaneously curious about AI (a Cognitive FX nationwide survey found more than 1 in 3 adults have used AI chatbots for some form of mental health support, often due to fear of judgment) and distrustful of AI in clinical care (multiple studies consistently show patients prefer human-in-the-loop, especially for anything beyond admin).

Your website now has to navigate both feelings at once.

The Three Categories of AI in a Therapy Practice

Before you can decide what to disclose, you need a clean mental model of what AI you're actually using. They are not the same thing, and they don't carry the same disclosure weight.

Category 1: Clinical AI (talks to clients about mental health)

Examples: an AI chatbot embedded on your homepage that triages, suggests coping strategies, or "supports" a visitor in crisis. White-labeled "AI therapy" tools. AI companion bots.

Doctive position: Don't.

In 2026, putting a clinical-facing AI chatbot on a licensed therapist's marketing website is the single highest-risk product decision available to you. The Brown University findings, the FTC review, and the documented safety incidents involving vulnerable users make this a clear professional-liability concern. Even a well-intentioned "supportive" bot can fail at crisis detection or cross a clinical line you didn't intend.

If a prospect needs immediate support, your site should route to a crisis line (988 in the US) and a human — never to an AI.

Category 2: Clinical-adjacent AI (handles your clinical work behind the scenes)

Examples: AI note-generation tools (built into Carepatron, SimplePractice, Heidi, Blueprint, etc.), session transcription, AI-assisted treatment planning, AI summarization of intake forms.

Doctive position: Use it carefully, disclose it explicitly.

Per APA ethical guidance, these tools must:

• Be HIPAA-compliant with a signed BAA

• Be reviewed by a human (you) before anything goes into the chart

• Be disclosed to clients in your informed consent

The website is one appropriate place to make that disclosure visible — usually in your Privacy Policy and on an "About Our Approach" or "Technology & Privacy" page. We'll show templates below.

Category 3: Administrative AI (never touches PHI or the clinical relationship)

Examples: AI used by you or your web vendor to draft marketing copy, generate website images, automate scheduling reminders that contain no PHI, run an audit on your site, manage billing reconciliation.

Doctive position: Use it freely. Disclose at your discretion.

This is the layer Doctive itself runs on for operations — and it never sees PHI. Disclosure here is optional, but a sentence in your Privacy Policy noting that you use automation and AI for administrative tasks is good practice for privacy-conscious clients.

What Clients Actually Want to See

Research on patient attitudes consistently shows three things:

1. Patients want human involvement. A 2024 study on AI conversational agents in mental health found most participants desired some human involvement in AI-driven care and were specifically concerned about AI being seen as a replacement for therapy. A meaningful subgroup said they were comfortable with AI for administrative tasks but not for care delivery.

   
   

2. Privacy is the top concern. Across multiple studies and the Reddit sentiment analyses, "where does my data go" is the dominant theme. Clients want to know if their words could end up training a model owned by a company they've never heard of.

3. They appreciate transparency over either silence or hype. Clinicians who openly say "I use [tool] for notes; here's how I protect your privacy" earn trust. Clinicians who quietly use AI tools without disclosure risk losing that trust in a single discovery moment.

This is the heart of the marketing question. The therapist who says "I use Carepatron's AI notes feature so I can be more present with you in session — your data is HIPAA-protected and never used to train external models" is doing both good ethics and good conversion. The therapist who says "My practice is AI-powered" is creating doubt about whether they're still doing the human work.

The Disclosure Decision Tree

Walk this in order. Stop at the first action.

Question 1: Do you have a clinical-facing chatbot on your site?

• Yes → Remove it. Replace with crisis-line links and a "Book a Consult" CTA.

• No → Continue.

Question 2: Do you use any AI tool that processes session content, client communications, or clinical notes? (This includes AI note generators, transcription, summarization, treatment-planning aids — whether built into your EHR or standalone.)

• Yes → Required: update your informed consent. Recommended: add a short, clear disclosure on your website's Privacy Policy and on an "About Our Approach" page. Templates below.

• No → Continue.

Question 3: Do you (or your web vendor) use AI for administrative tasks — marketing copy, image generation, scheduling automation, audits?

• Yes → Optional but recommended: a one-sentence Privacy Policy note disclosing administrative AI use.

• No → No disclosure needed.

Website Disclosure Templates

These are paste-ready starting points. Adapt to your actual tools and consult your licensing board's specific requirements.

Template A: Clinical-adjacent AI (for therapists using AI notes, transcription, etc.)

About Our Use of AI Our practice uses a HIPAA-compliant AI tool ([Tool Name]) to help with clinical documentation. Here's what that means in plain language: What it does: Helps your therapist draft session notes more efficiently, so they can spend more time present with you and less time typing after sessions. What it does not do: It does not make clinical decisions, replace your therapist's judgment, or talk to you directly. Your privacy: [Tool Name] is HIPAA-compliant, has signed a Business Associate Agreement with our practice, and does not use your information to train external AI models. Your therapist still reviews and edits every note. You are always treated by a human clinician. You can opt out. If you'd prefer your sessions not be documented using AI assistance, just let us know — we'll accommodate that. Questions? Bring them up with your therapist or email us at [practice email].

Template B: Administrative-only AI (optional but good practice)

A Note on Technology We use modern software and automation — including AI tools — to run administrative parts of our practice efficiently (such as marketing, scheduling reminders, and billing reconciliation). These tools never have access to your clinical records or session content. Your protected health information stays inside our HIPAA-compliant EHR, [Carepatron / SimplePractice / TherapyNotes]. We do not use AI in your clinical care. Your sessions, treatment, and clinical decisions are 100% led by your licensed therapist.

Template C: What to put on your homepage if you want to address it proactively

If you want to make your no-AI-in-clinical-care stance a competitive advantage (it increasingly is), a single sentence near your bio or "Approach" section works:

"Therapy here is human-led. I use modern tools to run the administrative side of my practice efficiently, but your sessions, your treatment plan, and our relationship are 100% me — a licensed [credential] working with you, not an algorithm."

What NOT to Put on Your Website

A few specific phrasings to avoid in 2026:

• ❌ "AI-powered therapy" — Reads as either misleading or genuinely concerning. The APA has formally asked the FTC to investigate AI products marketed as therapy.

• ❌ "AI-matched to your perfect therapist" — Used by some directories, but on a solo practice's site it suggests the clinical match itself is automated.

• ❌ "24/7 AI support between sessions" — Implies your AI is doing clinical work. If you mean "AI scheduling availability 24/7," say that.

• ❌ Embedded chatbots labeled "Chat with [Therapist Name]" that are actually AI. This is a direct misrepresentation under both APA transparency principles and FTC guidance.

• ❌ Silence when you do use clinical AI. A client who later discovers their notes were drafted by AI — and they were never told — has grounds for both an ethics complaint and a board complaint in many states.

• ❌ Stock AI imagery (glowing blue brains, robot heads) on a therapy site. Wrong cultural signal even if you mean it as decorative.

The Doctive Position (For Therapists Using Our Platform)

Doctive practices what we recommend:

• Wix attracts, Carepatron protects. Your marketing site is built with modern tools — including AI in our build process for things like content drafts, audits, and image generation — but the marketing site itself does not collect or process PHI.

• All PHI lives in Carepatron, which is HIPAA-compliant, signs BAAs, and is where any AI clinical-documentation features should run (with proper disclosure to your clients).

• No clinical chatbots, ever. We don't build them, we don't recommend them, and we won't add one to your site.

• We disclose our own AI use. Doctive uses AI for marketing automation, lead audits, and operations — never for clinical work and never with PHI. We're happy for you to see exactly where AI sits in our process.

The principle: AI is great for operations. Humans are necessary for relationships. Your website should make that distinction visible, not blurry.

Acceptance Criteria

Your AI positioning is healthy when:

• ✅ No clinical-facing chatbot exists on any page of your site

• ✅ Any clinical-adjacent AI tool you use is HIPAA-compliant with a signed BAA, disclosed in informed consent, and reviewed by you before anything goes into the chart

• ✅ Your Privacy Policy includes a clear, plain-language paragraph about your AI use (or non-use)

• ✅ Your homepage messaging makes it obvious that a licensed human leads the care

• ✅ A prospect could open your site cold and understand your AI stance in under 90 seconds

• ✅ Your stance can withstand a board complaint, an ethics audit, and a privacy-conscious client's questions

Next Actions

1. Run a free Doctive Audit on your therapist website — we'll flag any AI-related disclosure gaps, check your Privacy Policy against current APA/ACA guidance, and identify any chatbot or AI feature that should be removed. 48-hour turnaround, no commitment.

2. Get a personalized Wix Studio demo site with the right AI disclosure language built in, plus Carepatron booking — so you can see what a 2026-compliant therapy site actually looks like.

3. Explore the Doctive Growth Plan — $74/month or $777 one-time, all-in: custom domain, Wix Studio premium, Carepatron Advanced, Google Workspace, hosting, automations, updates, and support. No hidden charges. Wix attracts, Carepatron protects.

👉 Request your audit: hello@doctive.org | doctive.org

Sources & References

• American Psychological Association, Ethical Guidance for AI in the Professional Practice of Health Service Psychology (June 2025) (apa.org/topics/artificial-intelligence-machine-learning/ethical-guidance-ai-professional-practice)

• American Psychological Association, Health Advisory: Use of Generative AI Chatbots and Wellness Applications for Mental Health (November 2025)

• American Counseling Association, Recommendations for Practicing Counselors and Their Use of AI

• Brown University, Ethical Risks of ChatGPT as a Therapist (March 2026)

• APA Monitor, AI in the Therapist's Office: Uptake Increases, Caution Persists (March 2026)

• Patient Attitudes Toward AI Conversational Agents in Mental Health (PMC 11826059, 2024)

• Psychotherapists' Trust, Distrust, and Generative AI Practices in Psychotherapy, JMIR (April 2026)

• Cognitive FX, Survey: More Than 1 in 3 People Use AI Chatbots for Mental Health Support (January 2026)