Aug 17, 2026
The therapy room is no longer a purely human space. Ambient AI scribes listen to sessions and generate clinical notes. AI-powered chatbots offer emotional support between appointments. Large language models assist with treatment planning and documentation. And in 2026, these technologies are no longer optional curiosities—they are shaping how therapy is practiced, documented, and billed.
Yet the same features that make AI tools appealing—their ability to record, transcribe, analyze, and learn from clinical data—also create profound ethical and legal obligations. The question is no longer whether therapists should obtain informed consent for AI use, but how to do it properly, thoroughly, and defensibly.
As one leading ethics guideline states, "Maintain complete transparency by updating informed consent documents and discussing the specific use of AI tools directly with your clients". This article provides a comprehensive overview of the legal requirements, ethical standards, and practical steps for obtaining informed consent when using AI in therapy in 2026.
The Legal Landscape — New Laws Are Changing the Rules
2026 has seen a wave of state legislation regulating AI in mental health care. These laws establish clear, enforceable standards for informed consent that therapists cannot ignore.
State Legislation Requirements
Several states have enacted laws specifically addressing AI use in psychotherapy:
Colorado (HB 1195, effective August 12, 2026): Requires written informed consent before AI recording or transcription of therapeutic sessions. The law also prohibits AI from engaging in therapeutic communication without real-time professional involvement and bars AI-generated treatment recommendations without professional review.
Rhode Island (S2197, effective June 22, 2026): Mandates written informed consent before AI is used in recorded or transcribed therapeutic sessions. Patients must be informed in writing that AI will be used, told the specific purpose, and provide revocable consent.
Illinois, Nevada, Tennessee, and Utah: These states have also passed legislation restricting AI chatbots from being misrepresented as licensed behavioral health professionals and requiring disclosure to patients when AI is used in their care.
Maine: Effective January 1, 2026, Maine barred providers from using AI for therapeutic communications or treatment decisions and required patient consent before ambient listening tools may be used.
A Greenberg Traurig analysis notes that these laws are "designed to ensure that licensed clinicians – not AI tools – deliver psychotherapy, while permitting limited use of AI for operational support".
Key Legal Requirements Across Jurisdictions
Across these new laws, several common requirements emerge:
Written Informed Consent: Verbal consent is insufficient. Patients must provide written, documented consent before AI is used to record or transcribe sessions.
Specific Disclosure: Patients must be informed in writing about the specific purpose of AI use. Generic consent forms that simply mention "technology" or "digital tools" may not satisfy these requirements.
Revocability: Consent must be revocable, and refusal or revocation cannot result in denial of psychotherapy services.
Professional Oversight: AI tools may only be used for administrative or supplementary support, not independent clinical decision-making.
No AI Therapy: AI may not be marketed as, or function as, a substitute for licensed psychotherapy.
Professional Ethics — What APA, CAMFT, NASW, and AAMFT Now Require
Professional organizations have rapidly developed ethical guidance on AI use in therapy. A 2026 CE podcourse on "AI Ethics Codes for Therapists" examines how the CAMFT, APA, NASW, and AAMFT ethics codes now regulate AI in therapy.
The core ethical principles are not new; they are established principles applied to new technology. As one analysis notes, "Autonomy and informed consent, non-maleficence, beneficence, and justice are being applied to AI rather than reinvented".
APA Guidance
The American Psychological Association has released "Ethical Guidance for Professional Practice with Artificial Intelligence" in 2026. This guidance addresses critical considerations for:
Privacy and confidentiality
Clinical competence
Informed consent
APA's 2026 Chatbots and Mental Health Survey found that 77% of psychologists report their patients are using AI. More than a third say patients are using AI as an additional mental health provider. This widespread patient use makes informed consent not just an ethical obligation but a clinical necessity.
CAMFT's Three Obligations
The California Association of Marriage and Family Therapists separates three obligations that travel together:
Disclosure: Informing clients that AI is being used
Informed Consent: Obtaining explicit permission
Informed Decision Making: Ensuring clients understand the implications
Clients have the right to opt out.
ACA, AMHCA, and NBCC Guidance
Professional counseling organizations have issued position statements emphasizing that counselors must:
Disclose AI use to clients
Obtain informed consent
Safeguard client data
Ensure AI tools do not supplant the therapeutic relationship
The Consent Challenge — What Must Be Disclosed
Informed consent for AI use requires more than a checkbox. The 2026 guidance from multiple sources identifies specific elements that must be included.
Essential Disclosure Elements
According to emerging legal and ethical standards, therapists must disclose:
1. That AI Is Being Used
The patient must be explicitly told that AI tools are being used in connection with their care. This includes AI scribes, documentation tools, transcription services, and any AI-assisted analysis.
2. The Specific Purpose
Patients must be informed of the specific purpose of AI use. "We use AI for notes" is insufficient. Patients should understand exactly what the AI does with their session data.
3. How Data Is Handled
Patients must understand how their data is collected, stored, processed, and protected. This includes whether data is used to train AI models.
4. Risks and Limitations
Patients must be informed of the risks, including:
AI may produce inaccurate or fabricated information (hallucinations)
AI platforms may not offer the privacy protections of a clinical relationship
AI is not a crisis resource
AI may reflect algorithmic biases
5. The Option to Opt Out
Patients must understand they can refuse or revoke consent at any time without losing access to therapy.
The Separate Consent Issue
An emerging concern is that many digital mental health platforms use therapy transcripts to train AI models, often relying on lengthy terms of service to obtain assent. A 2026 npj Digital Medicine article argues that this "contractual assent is not informed consent".
The authors call for "separate and explicit opt-in consent to train on patient data and patient-led governance to prevent the recurrence of exploitation in medical innovation". This means that consent for AI-assisted documentation and consent for using data to train AI models should be separate, not bundled.
Practical Implementation — A Step-by-Step Guide
Step 1: Update Your Informed Consent Documents
Your informed consent forms must explicitly address AI use. This is now required by law in multiple states and recommended by professional ethics guidance.
Sample AI Consent Language:
"I understand that [Practice Name] uses AI-assisted documentation tools to support clinical record-keeping. These tools transcribe session content to generate progress notes, which are reviewed and approved by my therapist before becoming part of my permanent record. I understand that:
AI tools are used only for administrative and documentation support
All AI-generated content is reviewed by a licensed professional
My data is protected according to HIPAA and applicable state laws
AI outputs may be inaccurate and are not used for clinical decision-making
My data is not used to train AI models without my explicit separate consent
I may refuse or revoke this consent at any time without affecting my access to therapy"
Step 2: Discuss AI Use Directly with Clients
Simply including AI language in a consent form is insufficient. Therapists should discuss AI use directly with clients. This includes:
Explaining what the AI does and why you use it
Addressing any concerns or questions
Emphasizing that you remain responsible for all clinical decisions
Step 3: Verify Written Consent Before Recording
If you are using an AI scribe that records or transcribes sessions, you must obtain written informed consent before any AI recording or transcription occurs. This means:
The consent form must be signed before the first AI-recorded session
Consent cannot be retroactive
The specific purpose of AI use must be disclosed in writing
Step 4: Maintain Consent Records
Document that informed consent was obtained, including:
The date consent was given
The specific disclosures made
The patient's questions and your responses
Step 5: Review AI Outputs
Professional guidance consistently emphasizes that AI-generated clinical content must be reviewed and approved by a licensed clinician. Informed consent should reflect that the therapist remains responsible for all clinical decisions and documentation.
Special Considerations
When Patients Are Using AI Without Your Knowledge
APA's 2026 survey found that 77% of psychologists say their patients report using AI. Many patients use AI chatbots and wellness apps between sessions, often without disclosing this to their therapist.
APA recommends proactively asking patients about their AI use and creating room to explore what they are finding. This includes:
Normalizing that many people use AI for support
Exploring how the tool has been helpful or unhelpful
Discussing safety concerns, including that AI can give incorrect information and is not suitable for crises
Data Privacy and HIPAA
AI tools introduce significant privacy considerations. When selecting AI tools, therapists should:
Ensure the tool offers a Business Associate Agreement (BAA) under HIPAA
Verify where and how data are stored
Confirm that the vendor does not use client information to train their AI models
Understand that a signed BAA is not a guarantee of full HIPAA compliance
Training Data Consent
A critical emerging issue is whether patient data is used to train AI models. A 2026 article in npj Digital Medicine argues that "companies increasingly train AI models on therapy transcripts, often relying on terms of service that obscure data use to obtain assent".
Therapists should:
Understand whether their AI vendor uses client data for model training
Obtain separate, explicit opt-in consent for any such use
Be transparent with clients about how their data is used
Crisis Situations
Research indicates that patients are particularly concerned about AI use in crisis situations. Informed consent should address that AI is not a crisis resource and that patients should contact their therapist or emergency services directly in a crisis.
Enforcement and Consequences
State Enforcement
State laws carry enforcement mechanisms. Colorado's HB 1195 provides for professional discipline by licensing boards for violations. The Rhode Island law authorizes the Executive Office of Health and Human Services to investigate violations.
Professional Discipline
Violations of ethical standards regarding AI consent could result in:
Licensing board complaints
Professional discipline
Malpractice liability
Consumer Protection
Laws also restrict marketing that could mislead patients into believing AI is equivalent to licensed therapy.
FAQ
Do I need separate consent for AI use and for AI training data?
Yes. Emerging guidance recommends separate and explicit opt-in consent for using patient data to train AI models. Consent for AI-assisted documentation and consent for data training should be separate, not bundled.
Is verbal consent sufficient for AI use?
No. Multiple state laws now require written informed consent before AI recording or transcription of therapeutic sessions. Professional ethics also increasingly recommend written documentation.
What if my patient refuses AI consent?
Patients have the right to refuse or revoke consent without being denied psychotherapy services. You must have a workflow for providing therapy without AI tools for patients who opt out.
Does informed consent apply to AI chatbots patients use independently?
While you are not responsible for your patient's independent use of AI chatbots, APA recommends proactively asking patients about their AI use and discussing safety concerns. Some practices are adding questions about AI use to intake forms.
What should I do if my AI vendor uses patient data for training?
You must disclose this to patients and obtain explicit, separate consent. Many vendors now offer options to opt out of data training. Ensure you understand your vendor's data practices and communicate them clearly to patients.
How often should I update AI consent forms?
As laws and technologies evolve, you should review your AI consent forms at least annually or whenever you implement new AI tools. State laws enacted in 2026 are the first wave; more are expected.
References
If you’re ready to spend less time on documentation and more on therapy, get started with a free trial today
Not medical advice. For informational use only.
Outline
More articles
