Back to Ask the Experts

What Legal Risks Should Clinicians Consider Before Relying on AI Tools in Telehealth Practice?

Published on Nov 16, 2026

Question

What legal risks should clinicians consider before relying on AI tools in telehealth practice?

Answer

No AI vendor agreement currently guarantees the accuracy of the information the tool produces, even when marketing materials describe its capabilities favorably. This gap matters because the legal liability for any resulting error, incomplete information, or overly broad clinical statement falls on the clinician and their license, not on the AI vendor or the end user who happened to rely on the output. This risk extends to compliance work as well: using AI to draft telehealth policies or to research whether a particular legal requirement is in effect is unreliable, since AI models can generate outdated or simply incorrect information, and much of the detailed regulatory information clinicians need is not accessible to these tools in the first place.

A related concern involves AI-driven clinical flagging systems, such as tools that automatically flag any indication of depressive symptoms as a suicide risk. Over-flagging can create a "boy who cried wolf" problem that generates excessive documentation and dilutes attention to genuine risk, so any decision to disable such a flag should be restricted to the lead practitioner rather than administrative staff, and should be based on the clinician's own training and judgment rather than deferred to the software. The consistent theme is that just because AI can perform a task does not mean it should be trusted to do so without direct clinical oversight and independent verification.

  • Annemarie Kelly

    JD, LLM

Related Courses