Keep real information out
Do not enter real patient, workplace, classmate, family, or personal information into this learning environment or into an AI tool.
M
Metropolitan General Health System
PMHNP Clinical Learning Environment
New Provider Onboarding · Stop 5
Confidentiality, AI & Professional Expectations
Metropolitan General Health System is fictional, but the habits you practice here should reflect real clinical professionalism: protect information, reason from evidence, and use technology with care.
Six professional commitments
Use the environment—and AI—responsibly
Treat simulation records professionally
Use fictional charts with the same discretion, respect, and professional conduct expected in clinical practice. Do not post or casually share case content.
Use only the evidence available
AI cannot interview the patient, open an unreleased chart section, or create missing facts. Clearly distinguish documented evidence from assumptions.
Verify before you rely
Check AI-generated clinical claims, medication information, recommendations, calculations, and references against credible sources and the actual record.
Keep clinical judgment human
AI may help organize, compare, question, or communicate ideas when its use is permitted. You remain responsible for the reasoning and decisions you present.
Follow the directions for the activity
Expectations for whether, when, and how AI may be used can vary. Use the guidance provided in your course learning platform for each activity.
A practical boundary
AI can support a process. It cannot supply clinical evidence.
Useful support may include organizing ideas, identifying questions to investigate, comparing approaches, or improving patient-facing language—when that use is allowed.
- Use fictional, de-identified case information only.
- Check output against the chart and trusted sources.
- Identify uncertainty, limitations, and missing evidence.
- Revise or reject unsafe and unsupported suggestions.
- Entering protected or personally identifying information.
- Inventing symptoms, history, findings, or patient preferences.
- Accepting citations or recommendations without verification.
- Substituting generated output for independent judgment.