RINTARA INSIGHTS

Check Before Use: An AI Output Review Checklist

The TEPAT method for reviewing facts, assumptions, privacy and accountability before using an AI answer.

Learning outcome

After this module, you can review AI answers systematically before using them in work or decision-making.

The TEPAT method

1. Task: Does the answer genuinely solve the original task?
2. Evidence: Which facts, figures and sources require verification?
3. Privacy: Does the prompt or answer expose sensitive information?
4. Assumptions: What has the AI assumed without evidence?
5. Responsibility: Who will review and approve the final output?

Review prompt

Review the following AI answer as a critical reviewer.

Original task: [PASTE TASK]
AI answer: [PASTE ANSWER]

Identify:
1. Factual claims that require verification.
2. Unstated assumptions.
3. Information that does not serve the objective.
4. Privacy, bias or interpretation risks.
5. Recommended corrections.

Do not validate a claim without a source that can be checked.

10-minute exercise

Select an AI answer you have used before. Run the review prompt, check at least two claims against suitable sources, then rewrite the final output using your own corrections.

Module completion checklist

• I can distinguish facts, suggestions and assumptions.
• I do not enter confidential data into public platforms.
• I check numbers, names, dates and sources.
• I remain responsible for the final output.