It can, but a bot that asks about symptoms is collecting health information. That calls for a business associate agreement behind it, a narrow purpose and a hand-off to staff. For booking, a name, a phone number and a time are usually enough. Anything clinical belongs with a person, and the bot should not advise.
What changes when a bot asks about symptoms
Patients already type symptoms into chat boxes. A Pew Research Center survey from February 2026 found 20% of US adults ever use AI chatbots to get medical advice1, and a KFF poll fielded in early 2026 found 32% turned to an AI chatbot for health information in the past year2. The questions differ, so the figures do not compare, but a visitor may describe symptoms whether or not your bot asks, and the design has to plan for that.
In an MGMA poll of April 2025, 19% of medical group practices used some version of a chatbot or virtual assistant for patient communication3. Our test for a tool that must run inside a HIPAA-ready program is whether it can tell that a named person is a patient or is asking to become one4. A chat that takes names and symptoms meets it.
Ask only what the task needs
HHS's minimum necessary requirement asks covered entities to take reasonable steps to limit protected health information (PHI) to the minimum necessary. A chat built for the front desk can follow it task by task.
| Task | The bot may need | Leave to staff |
|---|---|---|
| A new patient asks for a visit | Name, callback number, preferred time, new or returning | Symptoms, history, any judgment of urgency |
| Hours, location, accepted insurance | Nothing about the person | Exceptions to the list |
| A refill or lab result request | Name and number, written down for staff | Looking up, giving or commenting on results |
| A visitor describes symptoms | A message to call 911 in an emergency and an offer to book | Every clinical decision |
Transcripts hold whatever was typed. Decide who may read them and how long they are kept, and log who opens them.
What the bot does when someone describes symptoms
Our site assistant answers questions about hours, insurance and booking and hands anything clinical to your staff. It should state only facts you supplied and say when it does not know. A British Columbia tribunal found an airline liable for negligent misrepresentation because its website chatbot told a customer he could apply for bereavement fares retroactively5. We use it as a design lesson; it is no rule for a US practice.
When a visitor does describe symptoms, the bot should not diagnose or reassure. It should say to call 911 for an emergency, offer a visit or a callback, and pass the message to a named person.
What some rules say about AI that talks about health
We are not lawyers. Read the rule yourself or ask your counsel how it applies to you.
California Health and Safety Code section 1339.75 requires a health facility, clinic, physician's office or group practice office that uses generative AI for patient communications about clinical information to include a disclaimer that the communication was AI-generated and clear instructions for reaching a human6. It excludes administrative matters such as appointment scheduling and billing6, and the disclaimer does not apply when a licensed provider reads and reviews the communication6.
Texas section 552.051(f) requires a provider of a health care service in which AI is used to give the disclosure no later than the date the service is first provided7. AMA policy from November 2024 says AI use should be disclosed at the start of an interaction8. We tell every visitor at the start that they are talking to an AI assistant.
People also ask
Can I use ChatGPT or Claude as a HIPAA-ready website chatbot?
Only through routes that come with a business associate agreement, such as the OpenAI API or the Claude API, and with the widget and host covered too. Patient details should not go into ChatGPT Business or Claude Team, which have no agreement available9.
Does the page a chatbot sits on change the HIPAA risk?
Yes. A symptom typed into a chat on a page with ad pixels or analytics can reach the tag vendor. HHS says a vendor that receives appointment details from a clinic's site is a business associate and needs an agreement10.
What does a HIPAA-ready chatbot cost?
Our HIPAA-ready Essentials program is $1,500 to set up and $699 a month per location4. The site assistant belongs to the Engineered plan at $2,999 a month plus a $3,000 setup, $1,400 on a 12-month term11. Usage is extra.
Can a chatbot give medical advice?
It should not. The Moffatt tribunal held an airline responsible for what its chatbot said5. A practice's bot is safest scoped to hours, insurance, booking and messages, with every clinical question recorded word for word and passed to a nurse or doctor.
Does a chatbot have to say it is an AI?
Some rules say so. Utah gives a safe harbor to a business whose generative AI says clearly, at the start and throughout, that it is an AI assistant12. California and Texas add disclosure duties for some health settings. We tell every visitor at the start that they are talking to an AI assistant.
Sources
- Pew Research Center. Americans and AI 2026: Chatbots, Smart Devices and Views on Impact. Pew Research Center, 2026. pewresearch.org/internet/2026/06/17/americans-and-ai-2026-chatbots-sma
- KFF. Poll: 1 in 3 Adults Are Turning to AI Chatbots for Health Information, Equaling the Share Who Use Social Media for Health. KFF, 2026. kff.org/health-information-trust/poll-1-in-3-adults-are-turning-to-ai-
- MGMA. Sizing up the market for AI chatbots and virtual assistants in medical practices in 2025. Medical Group Management Association (MGMA Stat), 2025-04-08. mgma.com/mgma-stat/sizing-up-the-market-for-ai-chatbots-virtual-assist
- Technical Marketing. HIPAA-ready programs. Technical Marketing, 2026-11. technicalmarketing.ai/hipaa/ Our own data.
- Christopher C. Rivers, Tribunal Member, Civil Resolution Tribunal (British Columbia). Moffatt v. Air Canada, 2024 BCCRT 149. Civil Resolution Tribunal of British Columbia, 2024-02-14. decisions.civilresolutionbc.ca/crt/crtd/en/525448/1/document.do
- California Legislature. California Health and Safety Code section 1339.75 (Chapter 2.13, Artificial Intelligence in Health Care Services). California Legislative Information (leginfo.legislature.ca.gov), 2024. leginfo.legislature.ca.gov/faces/codes_displaySection.xhtml?lawCode=HS
- Texas Legislature. H.B. No. 149 (enrolled), Texas Responsible Artificial Intelligence Governance Act. Texas Legislature Online (capitol.texas.gov), 2026-01-01. capitol.texas.gov/tlodocs/89R/billtext/html/HB00149F.htm
- American Medical Association. Augmented Intelligence Development, Deployment, and Use in Health Care (AMA policy). American Medical Association, 2024-11-12. ama-assn.org/system/files/ama-ai-principles.pdf
- Technical Marketing. Which software will sign a HIPAA BAA? 28 common tools, checked. Technical Marketing, 2026-11. technicalmarketing.ai/reports/hipaa-baa-availability/ Our own data.
- HHS Office for Civil Rights. Use of Online Tracking Technologies by HIPAA Covered Entities and Business Associates. U.S. Department of Health and Human Services, 2024-06-26. hhs.gov/hipaa/for-professionals/privacy/guidance/hipaa-online-tracking
- Technical Marketing. Pricing. Technical Marketing, 2026-11. technicalmarketing.ai/pricing/ Our own data.
- Utah Legislature. Utah Code 13-77-104 Safe harbor. Utah State Legislature (le.utah.gov), 2025-05-07. le.utah.gov/xcode/Title13/Chapter77/C13-77-S104_2025050720250507.html