This module develops the analytical skills needed to trace how information moves through an AI-assisted clinical operation from initial encounter capture through note generation, task and coordination communication, and final downstream record use. Learners will be able to apply a structured information flow analysis to a provided clinical operations scenario, identify where AI-generated content enters and exits each workflow stage, and document information flow gaps or handoff risks. Understanding the full information flow in AI-assisted clinical operations is essential for anyone responsible for ensuring that AI-generated content reaches the right destination accurately and that no critical information is lost or misrouted between workflow stages.

This module covers two foundational responsible AI skills for healthcare professionals working with Copilot tools: understanding where AI assistance appropriately stops and clinician judgment must take over, and evaluating AI-generated content for three specific risk categories—privacy risk, unsupported inference, and inappropriate disclosure—before that content is saved or sent. Learners will be able to apply a structured boundary framework to documentation and communication scenarios, and conduct a risk review of AI-generated content using defined evaluation criteria. These skills form the ethical and practical foundation for responsible governance of AI outputs in clinical environments.

This module develops the practical skill of applying verification checkpoints—structured review steps embedded at defined points in a routine workflow—to Microsoft 365 Copilot and Dragon Copilot outputs before those outputs are used, saved, or distributed in a healthcare setting. Learners will be able to design and apply a verification checkpoint sequence for a provided clinical workflow, implement each checkpoint using defined stewardship criteria, and monitor compliance with the checkpoint sequence over time. Verification checkpoints are the operational mechanism that translates responsible AI principles into consistent, repeatable practice—they are what makes governance real in day-to-day clinical operations.

In this project, learners produce a responsible AI review checklist for clinical use—a professional artifact that integrates clinical informatics understanding of how Copilot outputs interact with clinical systems, information flow analysis, responsible AI boundary application, privacy risk evaluation, and verification checkpoint design for a single realistic healthcare AI scenario. Learners who are responsible for reviewing or approving AI-generated content will be able to evaluate and document a complete responsible AI review approach for a bounded clinical Copilot workflow, producing a checklist that a clinical team could realistically adopt and use.