Using a Large Language Model to Identify Adolescent Patient Portal Account Access by Guardians

This diagnostic/prognostic study assesses the ability of a large language model (LLM) to detect guardian authorship of messages originating from adolescent patient portals.