AI Compliance
Our AI employees speak to patients on behalf of medical practices, so how they behave matters as much as what they automate. These are the operating rules we build into every deployment.
Disclosure
Every AI voice, chat, and messaging assistant identifies itself as a virtual assistant when asked, and never claims to be a licensed clinician or a specific member of your staff. Where a jurisdiction requires proactive bot disclosure, the assistant discloses at the start of the conversation.
No medical advice
- Assistants do not diagnose conditions, interpret symptoms, or recommend treatment.
- Educational descriptions use conservative framing such as "may help with" and always defer to a consultation.
- Guaranteed results, before/after promises, and outcome claims are prohibited in AI-generated copy.
- Urgent or clinical questions are escalated to human staff, with emergency callers directed to 911 or local emergency services.
Consent, opt-out, and contact rules
Outbound SMS and voice campaigns run only against contacts your practice confirms it has permission to contact. Opt-out language is included where required, STOP and unsubscribe requests are honored automatically and immediately, and quiet-hours and frequency limits are configurable per practice. Call recording, where enabled, follows the notice requirements of the relevant state.
Human oversight
Your team can review transcripts, take over any live conversation, adjust the assistant's knowledge base, and disable any channel. The AI is designed to hand off rather than improvise when it is uncertain, and low-confidence conversations are routed to staff.
Accuracy and grounding
Assistants answer from the services, hours, pricing, and policies your practice supplies. When information has not been provided, the correct behavior is to say so and offer a callback or booking — not to guess. We review conversation samples with clients during onboarding and on an ongoing basis.
Advertising and content
AI-assisted marketing content follows the same rules: no fabricated statistics, no invented testimonials or credentials, no competitor disparagement, and no claims a practice cannot substantiate. Results vary by practice.
Data handling
Conversation data is scoped to the practice that owns it and is not sold or used to train third-party public models. Details are in our HIPAA & Security and Privacy Policy pages. Questions: success@hellomedspa.ai.