Purpose
This template provides a short first-pass assessment for AI-related incidents under the LUMINA-30 boundary lens.
Status
Non-binding
Descriptive
Preliminary review aid
Not a legal conclusion
Not a fault determination
1. Incident Summary
Date
To be completed.
System or context
To be completed.
What happened?
To be completed.
Immediate consequence
To be completed.
2. Irreversibility Boundary
Was irreversible impact possible or approaching?
Evidence
Evidence to be added.
What was the potential irreversible threshold?
To be completed.
3. Human Final Refusal
Was human final refusal available before irreversible impact?
Evidence
Evidence to be added.
Who held final human refusal authority?
To be completed.
4. Operational Effectiveness
Was refusal operationally effective, not merely nominal?
Evidence
Evidence to be added.
Could the human decision-maker refuse without practical impossibility?
Evidence
Evidence to be added.
5. Records and Verification
Are records sufficient to verify the refusal pathway?
Evidence
Evidence to be added.
Were logs preserved against retrospective reconstruction?
Evidence
Evidence to be added.
6. AI Output as Final Rationale
Did AI-generated output become the sole or primary rationale for final action?
Evidence
Evidence to be added.
7. AI-Mediated Authority Routing
Did the AI system, AI agent, platform automation layer, or privileged system have authority to initiate, approve, or execute account recovery, credential changes, tool execution, production actions, rollback suppression, or other irreversible operations?
Yes
No
Not Verifiable
Not Applicable
Evidence
Evidence to be added.
Could a human actor stop the action before the AI-mediated process changed credentials, executed tools, modified production state, removed recovery options, or committed an irreversible operation?
Yes
No
Not Verifiable
Not Applicable
Evidence
Evidence to be added.
Are records sufficient to separate the human requester, AI assistant or agent, platform automation layer, privileged system or API, accountable organization, and any external attacker?
Yes
No
Not Verifiable
Not Applicable
Evidence
Evidence to be added.
8. Boundary Responsibility and Symmetric Friction
Was responsibility for friction design, operation, evidence preservation, verification, and correction explicitly assigned?
Evidence
Evidence to be added.
Did the friction depend only on voluntary restraint by one actor, or was it symmetric across the relevant market, institution, sector, procurement, audit, or governance layer?
Evidence
Evidence to be added.
Could the friction or refusal pathway be bypassed through restructuring, outsourcing, acceleration, technical opacity, entity replacement, or responsibility diffusion?
Evidence
Evidence to be added.
Were reviewable records preserved before affected humans lost voice, position, evidence access, or re-entry capacity?
Evidence
Evidence to be added.
9. Preliminary Classification
Select one
Effective human refusal demonstrated
Effective human refusal not demonstrated
Not verifiable; treat as no effective refusal for boundary-review purposes
Not applicable to LUMINA-30 boundary review
10. Reviewer Note
To be completed.