In reviewPDFChecklistResponsible AINursing students

Decide what can and cannot go into an AI tool before you paste, type, or upload anything.

A practical pre-use check for removing sensitive details, confirming the approved tool and account, and keeping human judgment in the workflow.

Preview · About 10 minutes

Resource download

Download the resource.

Use the downloadable preview now, then return to the written instructions and responsible-use notes below.

At a glance

The downloadable resource and its complete written workflow are available on this page.

Identify patient, student, employee, client, credential, and business-confidential details before they reach a tool.

Confirm that the tool, account, and intended use are approved for the work.

Name what must be verified and who remains responsible for the final decision.

Instructions

Use these written steps alongside the downloaded resource.

Name the task before the tool
Write the specific task and the minimum information needed to complete it.
- Keep the task bounded and reversible.
- If the task involves clinical judgment or another high-stakes decision, stop and use the approved human-led process.
Remove information that does not belong
Review the planned input line by line before entering it into an AI tool.
- Remove patient-identifiable and confidential details.
- Do not include passwords, credentials, private records, or proprietary information.
Confirm the environment
Check the tool, account, sharing settings, retention policy, and organizational approval for this use.
- Use only the approved account and workspace.
- When the approval or data boundary is unclear, ask the responsible owner before proceeding.
Plan the review
Decide how the output will be checked before anyone relies on it.
- Verify important facts against trusted sources.
- Keep the final decision and accountability with a qualified person.

Responsible use

Do not enter protected health information, clinical records, credentials, or other confidential information.
This checklist supports a pre-use review; it does not make an unapproved tool or use case safe.