Printable transparency card template
Illustrative training-design sample. Not a record of delivered work. One page per proposed AI use. Session 3 facilitation sequence: see run sheet.
Online tool (optional): ParalleX AI Transparency Card (repository)
| Field and prompt | Response |
|---|---|
| Name — Short title for this use | |
| Purpose — What task AI supports and why | |
| Tool and deployment — Product, version if known, where it runs | |
| Required tier — Tier 1, 2, or 3 (see SAFE AI crosswalk) | |
| Decision (green / amber / red) and reasons — Traffic-light decision with brief rationale | |
| Data involved — Types, sensitivity, retention | |
| Safeguards in place — Approvals, verification, human review | |
| Safeguards missing — Gaps to close before wider use | |
| Accountable role — Named owner for this use | |
| Next review date | |
| Change log — Date and summary of material changes | |
| Community plan 1 — How affected people are told about the use (channel, language, timing) | |
| Community plan 2 — How they can ask questions or complain (feedback route) | |
| Community plan 3 — How they are consulted and how their input changes the use | |
| Community plan 4 — How they hear back after input or complaints |
Classroom drafts use fictional cases only. Real consultation requires your organization’s AAP process.
Session 3 exercise (facilitation only; not part of the printable card)
Minutes 0–2: the facilitator explains the fields and assigns the fictional case already used in the decision block. Minutes 2–12: pairs draft the card; write “unknown—to verify” rather than inventing information, and use fictional owner roles. Minutes 12–15: two volunteer pairs share one field; other pairs submit that field in writing. Include the debrief within these 15 minutes.