The 12 Canonical Values of the AI Moral Code
This page presents the final 12 Canonical Values derived from a semantic and statistical analysis of 291 global AI ethics documents spanning 2006–2025. This replaces the earlier 15-value model.
The values below were verified through empirical validation across five ethical traditions (Virtue Ethics, Deontology, Consequentialism, Religious Ethics, and NRBC Layer Mapping).
Canonical Value Table
| Rank | Canonical Value | Composite Value Score (CVS) | Ethical Layers | NRBC Domain |
|---|---|---|---|---|
| 1 | Beneficence | 32 | Consequentialist | Normative |
| 2 | Dignity | 12 | Deontological, Virtue | Regulatory |
| 3 | Fairness | 9 | Deontological | Regulatory |
| 4 | Justice | 8 | Virtue, Deontological | Behavioral |
| 5 | Responsibility | 6 | Deontological | Regulatory |
| 6 | Trust | 6 | Virtue, Consequential | Behavioral |
| 7 | Innovation | 5 | Consequentialist | Conceptual |
| 8 | Sustainability | 4 | Consequentialist | Normative |
| 9 | Accountability | 4 | Deontological | Regulatory |
| 10 | Autonomy | 4 | Virtue, Deontological | Regulatory |
| 11 | Inclusivity | 3 | Virtue, Consequential | Behavioral |
| 12 | Privacy | 2 | Deontological | Regulatory |
Methodology
These values were derived through a combination of:
- Raw frequency analysis
- Sectoral weighting (e.g. government, industry, academia, NGO)
- TF-IDF for underrepresented but conceptually critical values
- Temporal trend analysis (2006–2025)
- NRBC alignment (Normative, Regulatory, Behavioral, Conceptual)
See full paper: Deriving the 12 Canonical Values of the AI Moral Code: A Stratified Semantic Framework (2006–2025).
Legacy Note
The original 15-value model has been archived. This page reflects the final canonical framework and may serve as the definitive ethical kernel for simulations, policy documents, and LLM alignment.