Iowa City-based University of Iowa Health Care is warning employees against the use of ChatGPT for its potential to violate HIPAA, per Becker’s.
In an Oct. 26 notice, the health system said the artificial intelligence chatbot is not HIPAA-compliant and thus, to be used, tools like it must undergo a “proper security review, contracting, and business associate agreement.”
“[Entering] patient information into an AI system could result in a HIPAA violation,” the warning read. “For example, using ChatGPT to draft a patient letter or using an unapproved AI transcription service requires sharing of (patient health information) with the application. Beware of these types of situations.”
