# [Agency Name] Artificial Intelligence Acceptable-Use Policy (MODEL TEMPLATE)
*This template is not legal advice. Review it with agency counsel, your privacy officer and the medical director before adopting it.*
1. **Purpose.** Allow AI tools that improve documentation, education and operations while protecting patients and PHI.
2. **Scope.** All employees, volunteers, students and contractors.
3. **Approved tools.** Only tools on the Approved AI List maintained by [role]. Tools that process PHI must have a signed BAA.
4. **PHI rules.** Do not enter PHI into any non-approved tool. De-identify data under the HIPAA Safe Harbor standard when it is used for training or templates.
5. **Clinical use.** AI does not replace protocols, medical direction or clinical judgment. AI must not be used for real-time treatment decisions unless the tool is approved by the medical director.
6. **Documentation.** The provider reviews, edits and attests to all AI-assisted ePCR content before signing.
7. **Verification.** Users verify all AI-generated facts, doses and citations.
8. **Transparency.** Disclose AI assistance where agency policy, law or a court requires it.
9. **Training.** Users complete AI and privacy training before use.
10. **Incidents.** Report suspected PHI exposure or harmful AI output to [privacy officer] within [24h].
11. **Enforcement and review.** Violations follow the disciplinary policy. This policy is reviewed annually.