A Proof System with Causal Labels (Part II): checking Counterfactual Fairness

2025-07-19

AI Summary

This paper addresses the problem of verifying counterfactual fairness for probabilistic classifiers. We propose a formal method based on typed natural deduction, extending the TNDPQ calculus with causally annotated structural conditions that embed structural causal models and probabilistic reasoning into the type system. This enables rigorous logical characterization of the counterfactual proposition: “Would the decision remain unchanged if a sensitive attribute were altered?” The resulting labeled proof system supports automated derivation and formally verifiable fairness certification, overcoming key limitations of traditional statistical fairness definitions—namely, their lack of causal semantics and formal provability. Empirical evaluation demonstrates that our framework effectively detects latent counterfactual unfairness in black-box classifiers, providing an interpretable and formally grounded mechanism for fairness assurance in trustworthy AI systems. (149 words)

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Ethics, Bias, and Fairness
• Philosophy and Ethics of AI: Bias, Fairness & Equity

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems

Abstract

In this article we propose an extension to the typed natural deduction calculus TNDPQ to model verification of counterfactual fairness in probabilistic classifiers. This is obtained formulating specific structural conditions for causal labels and checking that evaluation is robust under their variation.

Problem

• Extending TNDPQ calculus to verify counterfactual fairness
• Modeling structural conditions for causal labels variation
• Checking robustness of probabilistic classifier evaluations

Innovation

• Extends TNDPQ calculus for fairness verification
• Formulates structural conditions for causal labels
• Checks evaluation robustness under label variation