# AI Source Integrity Field Guide Workbook

## Setup
1. Name the workflow, document, decision, or learning activity under review.
2. Assign an accountable reviewer with authority to accept, revise, or reject the AI-assisted work.
3. Record the model, prompt, date, and source set before reviewing conclusions.

## Review Questions
- Source categories: what evidence proves this step was completed, and where is that evidence stored?
- Verification workflow: what evidence proves this step was completed, and where is that evidence stored?
- Citation repair: what evidence proves this step was completed, and where is that evidence stored?
- Red-flag language: what evidence proves this step was completed, and where is that evidence stored?
- Decision tree: what evidence proves this step was completed, and where is that evidence stored?

## Acceptance Record
| Item | Evidence Location | Reviewer | Decision | Follow-up |
| --- | --- | --- | --- | --- |
| 1 |  |  | Accept / Revise / Reject |  |
| 2 |  |  | Accept / Revise / Reject |  |
| 3 |  |  | Accept / Revise / Reject |  |

## Disclosure Note
This material supports accountable AI use by documenting human review, source verification, and governance decisions. It is informational and does not replace professional legal or regulatory advice.