Rhode Island Enacts AI Recording Consent Laws for Health Care

Rhode Island Enacts AI Recording Consent Laws for Health Care
Gov. Dan McKee signed three Rhode Island AI laws on June 22, 2026: a healthcare AI-scribe notification act (S2570 SUB A/H7538), a therapy-AI consent law (S2197 SUB A/H7349 SUB A), and an AI companion chatbot safety law (S2195 SUB A/H7350 SUB A), layering new consent duties atop the state's one-party consent recording statute.
Information last verified on July 4, 2026. This is a developing story; we update it as the record changes.
Jurisdiction scope: This article addresses Rhode Island state law only: three 2026 session bills signed by Gov. Dan McKee on June 22, 2026, and how they interact with Rhode Island's existing one-party consent recording statute, R.I. Gen. Laws Section 11-35-21. It does not address other states' AI-in-healthcare or AI-in-therapy laws. For a similar healthcare AI-disclosure law in another state, see Louisiana's AI medical-visit disclosure law under HB 475; for a comparable AI-therapy restriction, see Vermont's AI mental health therapy law under H.816.
What Happened
Rhode Island Gov. Dan McKee signed a three-bill package of AI laws on June 22, 2026. The first, S2570 SUB A (companion bill H7538), titled the Use of Artificial Intelligence by Healthcare Providers Notification Act, requires a healthcare provider or facility using AI to document, or "memorialize," a patient visit to notify the patient of that use and to review the AI-generated documentation for accuracy afterward. The bill was lead-sponsored by Sen. Pamela Lauria and Rep. Teresa Tanzi. The second, S2197 SUB A (companion bill H7349 SUB A), an oversight act for AI in mental health care, restricts AI's role in therapy and psychotherapy sessions. The third, S2195 SUB A (companion bill H7350 SUB A), regulates AI companion chatbots, requiring self-harm safeguards and a disclosure that the chatbot is not human.
Rep. Tanzi described the intent behind the healthcare notification bill in a statement following the signing:
"AI scribes and similar tools have the potential to decrease the documentation burden for medical providers and improve the quality of visits for patients. But as with any rapidly expanding new technology, particularly in a sensitive field like healthcare, it is important to protect patients and transparently disclose when AI scribes are being used."

What the Law Actually Says
Rhode Island is a one-party consent state for recording: under R.I. Gen. Laws Section 11-35-21, a person who is a party to a wire, electronic, or oral communication, or who has the prior consent of one party to it, may lawfully record that communication, so long as the recording is not made for a criminal, tortious, or other injurious purpose. See our Rhode Island recording laws guide for the full statute and its exceptions. That baseline rule does not, by itself, require anyone to disclose that AI, rather than a human, is doing the recording or transcription. The three 2026 laws add that disclosure and consent layer for specific, sensitive contexts.
Healthcare AI documentation (S2570 SUB A/H7538). The notification act applies broadly to healthcare providers, defined to include physicians, physician assistants, dentists, registered and licensed practical nurses, advanced practice registered nurses, and other professionals licensed by the state health department, as well as healthcare facilities. Reporting on the bill describes its core mechanic as requiring the provider to disclose AI-scribe use to the patient before or during the visit and to let the patient decline it, then to review the AI-generated notes for accuracy after the visit ends. This sits alongside, and does not replace, Rhode Island's existing medical recording rules; see our Rhode Island medical recording laws page on patient rights and HIPAA. The federal HIPAA Privacy and Security Rules at 45 C.F.R. Part 164 continue to govern how any AI-generated clinical documentation is stored, secured, and disclosed as protected health information, regardless of who or what created it.
AI in therapy (S2197 SUB A/H7349 SUB A). This act, titled an oversight act for artificial intelligence technology in mental health care, prohibits any individual or business from providing, advertising, or otherwise offering therapy or psychotherapy services in Rhode Island unless those services are delivered by a licensed professional. It separately prohibits AI from assisting in supplementary support or therapeutic communication in a therapy or psychotherapy session where that session is recorded or transcribed, unless the patient is informed in writing and provides consent. The act also states that a licensed provider may not use AI to make independent therapeutic decisions or to determine therapeutic recommendations or treatment plans. In practice, this stacks a written-consent requirement on top of the one-party consent baseline for the specific case of AI recording or transcribing a therapy session, a stricter rule than section 11-35-21 alone would require.
AI companion chatbots (S2195 SUB A/H7350 SUB A). This act, covering AI companion models, requires an operator to build in a protocol addressing expressions of suicidal ideation, self-harm, or potential physical harm to others by a user, and to respond by referring the user to crisis service providers such as a suicide hotline or crisis text line. It also requires the chatbot to notify users that it is not human and does not have human emotions. Operators must file annual reports with the Rhode Island attorney general on safety-protocol activations, beginning July 1, 2027, and the act takes effect January 1, 2027. Violations carry civil penalties of up to $15,000 per day, directed to suicide prevention programs; amendments during the bill's passage removed an earlier private right of action and a financial-harm crisis-protocol trigger.
Analysis: Why This Matters
The following is analysis from the Recording Law Editorial Team.
Rhode Island's package treats AI-mediated recording and documentation as functionally different from a human doing the same task, at least in two high-sensitivity settings: the exam room and the therapy session. Section 11-35-21 has long allowed a party to a conversation, including a healthcare provider, to record it with only that party's own consent. The new healthcare notification act does not disturb that baseline consent rule; it adds a disclosure duty specifically tied to AI tools, which suggests the legislature viewed the delegation of documentation to an AI system, rather than the recording itself, as the feature that warranted a distinct notice requirement.
The therapy-AI law goes further than disclosure alone. By requiring written consent before AI can assist in or record a psychotherapy session, and by reserving therapeutic decision-making and treatment planning to licensed humans, the statute draws a line between AI as a documentation or support tool and AI as a substitute clinician. That distinction tracks a broader pattern this year across states regulating AI in mental health care, including Vermont's H.816, which similarly restricts AI's role in therapy.
The companion chatbot law addresses a separate risk category: a user in emotional distress interacting with a system that is not a licensed provider and not human at all. Requiring crisis-service referral protocols and a not-human disclosure, backed by per-day civil penalties, positions the statute as a consumer-safety measure layered onto, rather than replacing, existing mental-health licensing law.
How This Affects You
Patients in Rhode Island who see a healthcare provider that uses an AI scribe should, under the new notification act, be told about that use and given the opportunity to decline it; providers are also expected to review AI-generated visit notes for accuracy. Patients considering AI-assisted therapy tools, or apps that record or transcribe therapy sessions, should look for written consent documentation and confirm that the service is delivered or supervised by a licensed mental health professional, since Rhode Island now restricts who may offer therapy or psychotherapy and how AI may be used within it. Users of AI companion or chatbot apps in Rhode Island should expect, once the January 1, 2027 effective date arrives, a disclosure that the chatbot is not human, and, if they express thoughts of self-harm, a referral to crisis resources such as a suicide hotline or crisis text line. None of these new consent layers change the state's underlying one-party consent rule for ordinary recordings; that rule, and its exceptions for criminal, tortious, or injurious purposes, still applies independently of whether AI is involved.
This is general legal information, not legal advice. It covers Rhode Island state law as reflected in sources verified on July 4, 2026. Laws change and this is a developing story; consult a lawyer licensed in Rhode Island about your specific situation.
Sources
- Rhode Island S2570 SUB A / H7538, Use of Artificial Intelligence by Healthcare Providers Notification Act (bill text)
- Rhode Island H7349 SUB A / S2197 SUB A, Oversight of Artificial Intelligence Technology in Mental Health Care Act (bill text)
- Rhode Island S2195 SUB A / H7350 SUB A, Artificial Intelligence Companion Models Act (bill text)
- R.I. Gen. Laws Section 11-35-21, Unauthorized interception, disclosure, or use of wire, electronic, or oral communication
- 45 C.F.R. Part 164, HIPAA Security and Privacy Rules (eCFR)
Related articles
- Rhode Island Recording Laws: One-Party Consent Rules
- Rhode Island Medical Recording Laws: Patient Rights and HIPAA Rules
- Rhode Island AI Laws and Regulation
- Vermont Enacts H.816 Restricting AI Mental Health Therapy
- Louisiana HB 475: Disclosure Required Before AI-Recorded Medical Visits
Last updated: 2026-07-04. This is a developing story; details verified as of 2026-07-04.
Frequently Asked Questions
What are Rhode Island's new AI recording and consent laws?
On June 22, 2026, Gov. Dan McKee signed three bills: S2570 SUB A/H7538 requiring patient notice before a healthcare provider uses AI to document a visit, S2197 SUB A/H7349 SUB A requiring written patient consent before AI assists in or records a therapy session, and S2195 SUB A/H7350 SUB A requiring AI companion chatbots to include self-harm safeguards and disclose they are not human.
Does a Rhode Island doctor have to tell me if they use an AI scribe?
Under S2570 SUB A/H7538, the Use of Artificial Intelligence by Healthcare Providers Notification Act, a healthcare provider or facility using AI to document a patient visit must notify the patient of that use and review the AI-generated documentation for accuracy afterward.
Can I opt out of AI documentation during a medical visit in Rhode Island?
Reporting on S2570 SUB A/H7538 describes the notification requirement as including the ability for a patient to decline AI-scribe use during the visit, though the statute's enforcement mechanism was not detailed in the sources reviewed for this article.
Can a therapist in Rhode Island use AI to conduct psychotherapy?
No. S2197 SUB A/H7349 SUB A bars anyone from providing, advertising, or offering therapy or psychotherapy in Rhode Island unless it is delivered by a licensed professional, and a licensed provider may not use AI to make independent therapeutic decisions or set treatment plans.
Is it legal for an AI tool to record or transcribe my therapy session in Rhode Island without my consent?
Under S2197 SUB A/H7349 SUB A, AI may not assist in or record/transcribe a therapy or psychotherapy session unless the patient is informed in writing and consents, a stricter requirement than Rhode Island's general one-party consent recording rule under R.I. Gen. Laws Section 11-35-21.
What do Rhode Island's new AI companion chatbot rules require?
S2195 SUB A/H7350 SUB A requires chatbot operators to include protocols that detect expressions of suicidal ideation or self-harm and refer the user to crisis services, and to disclose that the chatbot is not human. It takes effect January 1, 2027, with civil penalties of up to $15,000 per day for violations.
How does this relate to Rhode Island's one-party consent recording law?
Rhode Island remains a one-party consent state under R.I. Gen. Laws Section 11-35-21: a party to a conversation may record it with only their own consent. The 2026 AI laws add disclosure or written-consent requirements on top of that baseline for AI-mediated documentation in healthcare, therapy, and companion chatbot contexts; they do not repeal the underlying one-party consent rule.
When do Rhode Island's 2026 AI consent laws take effect?
Gov. McKee signed all three bills on June 22, 2026. The AI companion chatbot law, S2195 SUB A/H7350 SUB A, takes effect January 1, 2027, with annual attorney general reporting starting July 1, 2027. Reported summaries of the healthcare notification act, S2570 SUB A/H7538, describe it taking effect upon passage; a specific effective-date clause for the therapy-AI law, S2197 SUB A/H7349 SUB A, was not confirmed in the sources reviewed for this article.
Sources and References
- Rhode Island S2570 SUB A / H7538, Use of Artificial Intelligence by Healthcare Providers Notification Act (bill text)(webserver.rilegislature.gov).gov
- Rhode Island H7349 SUB A / S2197 SUB A, Oversight of Artificial Intelligence Technology in Mental Health Care Act (bill text)(webserver.rilegislature.gov).gov
- Rhode Island S2195 SUB A / H7350 SUB A, Artificial Intelligence Companion Models Act (bill text)(webserver.rilegislature.gov).gov
- R.I. Gen. Laws Section 11-35-21, Unauthorized interception, disclosure, or use of wire, electronic, or oral communication(webserver.rilegislature.gov).gov
- 45 C.F.R. Part 164, HIPAA Security and Privacy Rules (eCFR)(ecfr.gov).gov