Index

MEG Case Studies

Welcome to the MEG (Minimal Ethical Governance) Case Studies Index. This compendium serves as an operational and technical-legal testbed for the MEG framework (incorporating both the technical specifications of MEG 1 and the legal liability models of MEG 2).

To bridge the "Accountability Gap" in modern and Agentic AI systems, we subject real-world failures, catastrophic model drifts, and systemic historical incidents to rigorous normative evaluations under the MEG protocol, co-authored and structured in collaborative symbiosis with advanced generative AI assistants.

The index below contains case studies:

Cases 001 – 010: Modern and complex Agentic AI failures, platform security exploits, and commercial liability developments from 2024 to 2026 (including Replit, PocketOS/Cursor, and the Munich Re/HSB actuarial integrations).
Cases 011 – 110: A comprehensive compendium of historical AI and algorithmic incidents spanning facial recognition abuses, automated benefit fraud rejections, conversational harms, and predictive policing failures. Each historical case includes a dedicated analysis of "How MEG 1 would have acted" to prevent or mitigate the harm.

New cases (N-series) are documented after the framework publication and are maintained as a living annex. Unlike cases 001–110, they are not yet included in the archived Zenodo release and carry no persistent DOI; each is labelled Grounded, Probed or Analogical to indicate evidentiary status.

N001 N002 N003 N004 N005 N006 N007 N008 N009 N010 N011 N012 N013 N014 N015 N016 N017 N018 N019

For detailed technical schemas, persistent DOIs on Zenodo, and full framework specifications, please consult our primary documentation library:

Primary Documentation & Resources:
• Specification Library: meg-initiative.org/library/
• Zenodo Technical Specs (MEG 1): doi.org/10.5281/zenodo.21280675
• Zenodo Legal Framework (MEG 2): doi.org/10.5281/zenodo.21280679
• SSRN Academic Preprint (MEG 1): ssrn.com/abstract=7110079
• SSRN Academic Preprint (MEG 2): ssrn.com/abstract=7110178

Select any case below to see exactly how MEG 1 + MEG 2 would have prevented, mitigated, or allocated liability for the incident.