Moral Entropy: Auditing Bias and Uncertainty in Moral Judgment
Abstract
Most work in computational ethics treats annotator disagreement on moral content as noise to be voted away, collapsed into majority vote or the more permissive any-annotator rule the moment a single annotator flags an item. We argue this uncertainty should instead be modeled and learned from. We introduce Moral Entropy, a Bayesian framework that keeps a full posterior over the true label and decomposes its entropy into aleatoric uncertainty (irreducible disagreement about the moral content) and epistemic uncertainty (from insufficient or noisy annotation) -- and lets any heuristic consensus rule be audited against a calibrated ground truth via entropy methods such as cross-entropy/KL, Brier score, and expected calibration error. Across three corpora and fifteen discourse domains, auditing the standard aggregation rules against this posterior reveals bias that no current pipeline reports: the any-annotator rule disagrees with the calibrated posterior on roughly 30% of items -- pooled, almost entirely false positives, though the errors invert at the foundation level (19.9%/38.9% mean FPR/FNR on MFTC) -- while the stricter majority and two-vote rules miss 63-83% of true positives.
Related
- Efficient Active Auditing of Multi-Group Fairness with Bias Probes
- Bias-Corrected Estimators for Joint Entropy, Conditional Entropy, and Mutual Information in the Discrete Bivariate Case: Theory and an Application to Motor Insurance
- When Timing Uncertainty Limits Spectral Accuracy: Bias, Variance, and Frequency-Dependent Attenuation
- A Hierarchy of Entropy-Shapley Games for Multivariate Predictive Uncertainty
- Encoding Designers’ Aesthetic Judgment into Fast Generative Interior Design: Expert-Calibrated Reward Alignment with ARIA-Diff
BibTeX
@misc{skorski2026moral,
title = {{Moral Entropy: Auditing Bias and Uncertainty in Moral Judgment}},
author = {Maciej Skorski},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.21992},
eprint = {2609.21992},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.21992},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Moral Entropy: Auditing Bias and Uncertainty in Moral Judgment
AU - Maciej Skorski
PY - 2026
DA - 2026/09/18
PB - arXiv
DO - 10.48550/arxiv.2609.21992
UR - https://arxiv.org/abs/2609.21992
AB - Most work in computational ethics treats annotator disagreement on moral content as noise to be voted away, collapsed into majority vote or the more permissive any-annotator rule the moment a single annotator flags an item. We argue this uncertainty should instead be modeled and learned from. We introduce Moral Entropy, a Bayesian framework that keeps a full posterior over the true label and decomposes its entropy into aleatoric uncertainty (irreducible disagreement about the moral content) and epistemic uncertainty (from insufficient or noisy annotation) -- and lets any heuristic consensus rule be audited against a calibrated ground truth via entropy methods such as cross-entropy/KL, Brier score, and expected calibration error. Across three corpora and fifteen discourse domains, auditing the standard aggregation rules against this posterior reveals bias that no current pipeline reports: the any-annotator rule disagrees with the calibrated posterior on roughly 30% of items -- pooled, almost entirely false positives, though the errors invert at the foundation level (19.9%/38.9% mean FPR/FNR on MFTC) -- while the stricter majority and two-vote rules miss 63-83% of true positives.
ER -