AI Forensic Readiness v0.1: A Working Model for Investigating Consequential Actions Performed or Influenced by Autonomous AI Systems
Description
AI Forensic Readiness v0.1 proposes a working practitioner model for reconstructing, attributing, containing, validating, and recovering from consequential actions performed or influenced by autonomous AI systems.
The discussion draft introduces a six-question investigation model, a minimum viable evidence profile, the AI Incident Reconstruction Graph, dependency-aware containment, verifiable revocation, reversibility classes, and a six-level readiness maturity model.
This release is intended for controlled testing and public review and is not presented as a formal standard.
Public review is open through October 17, 2026:
https://github.com/Scoston/ai-forensic-readiness/issues/1
Licensing: specification prose, explanatory materials, assessments, and diagrams are licensed under CC BY 4.0. Schemas, examples, and code are licensed under Apache License 2.0.
Files
AI_Forensic_Readiness_v0.1.pdf
Additional details
Related works
- Is identical to
- Software: https://github.com/Scoston/ai-forensic-readiness/releases/tag/v0.1.0-draft (URL)
- Is version of
- Software: https://github.com/Scoston/ai-forensic-readiness (URL)