Beyond Verification: Abductive Explanations for Post-AI Assessment of Privacy Leakage

TL;DR

This paper introduces a formal abductive reasoning framework for auditing privacy leakage in AI models, providing interpretable explanations and privacy guarantees, demonstrated on credit scoring data.

Contribution

It proposes a novel abductive explanation-based framework for privacy auditing that formalizes both individual and system-level leakage with human-understandable outputs.

Findings

Abductive explanations can identify sensitive information disclosures.

The framework offers privacy guarantees while maintaining interpretability.

Experimental results on the German Credit Dataset illustrate the approach's effectiveness.

Abstract

Privacy leakage in AI-based decision processes poses significant risks, particularly when sensitive information can be inferred. We propose a formal framework to audit privacy leakage using abductive explanations, which identifies minimal sufficient evidence justifying model decisions and determines whether sensitive information disclosed. Our framework formalizes both individual and system-level leakage, introducing the notion of Potentially Applicable Explanations (PAE) to identify individuals whose outcomes can shield those with sensitive features. This approach provides rigorous privacy guarantees while producing human understandable explanations, a key requirement for auditing tools. Experimental evaluation on the German Credit Dataset illustrates how the importance of sensitive literal in the model decision process affects privacy leakage. Despite computational challenges and simplifying assumptions, our results demonstrate that abductive reasoning enables interpretable privacy auditing, offering a practical pathway to reconcile transparency, model interpretability, and privacy preserving in AI decision-making.