AI use-case risk screen
Choose whether an AI use case can proceed, needs redesign or belongs with a qualified provider.
Done when: The affected decisions, data boundaries, human authority and remaining risks are explicit.
Enable JavaScript for on-page entry and smart printing, or print a blank copy from your browser and complete it offline.
People and decisions affected
Describe who could be affected and the decision the system influences. Distinguish a draft suggestion from an action taken automatically.
Data categories and exclusions
Identify whether health, financial, employment, identity or other sensitive data could enter. Record categories only, never real records.
Failure severity
Describe a plausible wrong answer, missed action or data disclosure and the harm it could cause.
Human review and authority
Name the role that may approve consequential actions and the evidence they must see first.
Controls and remaining gaps
List access limits, test cases, monitoring, retention and shutdown controls. Keep unresolved risks visible.
Proceed, redesign or stop
Record the decision, rationale, accountable approver and next review trigger. This worksheet is not a safety certification.
Reference links and review boundary
Use the current issuer, payer, regulator, platform or qualified professional for your exact activity and location. These links informed the planning questions; they do not make this an official form, full compliance review or permission to operate.
Healthcare businesses are welcome for business education. PublicProof does not receive, store or process patient information. A completed worksheet does not establish HIPAA compliance, transport qualification, professional authority, account approval, a credential, reimbursement or revenue.
Worksheet updated October 20, 2026. Verify changing requirements before relying on them.