# Fairness of Explanations in Artificial Intelligence (AI): A Unifying Framework, Axioms, and Future Direction toward Responsible AI

arXiv: https://arxiv.org/abs/2605.09852 · HTML: https://mlsanity.com/paper/2605.09852/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 27 of its 227 references that are in the map.
- First posted on arXiv 11 May 2026.

## Builds on (most cited first, 27 of 27)

- [How Powerful are Graph Neural Networks?](https://mlsanity.com/paper/1810.00826/) (2018)
- [Evaluating and Aggregating Feature-based Model Explanations](https://mlsanity.com/paper/2005.00631/) (2020)
- [Aequitas: A Bias and Fairness Audit Toolkit](https://mlsanity.com/paper/1811.05577/) (2018)
- [Logical Neural Networks](https://mlsanity.com/paper/2006.13155/) (2020)
- [General Pitfalls of Model-Agnostic Interpretation Methods for Machine Learning Models](https://mlsanity.com/paper/2007.04131/) (2020)
- [Equalizing Recourse across Groups](https://mlsanity.com/paper/1909.03166/) (2019)
- [Metric Learning for Individual Fairness](https://mlsanity.com/paper/1906.00250/) (2019)
- [On the relation between accuracy and fairness in binary classification](https://mlsanity.com/paper/1505.05723/) (2015)
- [How model accuracy and explanation fidelity influence user trust](https://mlsanity.com/paper/1907.12652/) (2019)
- [Explainability for fair machine learning](https://mlsanity.com/paper/2010.07389/) (2020)
- [Auditing and Achieving Intersectional Fairness in Classification Problems](https://mlsanity.com/paper/1911.01468/) (2019)
- [Robust Semantic Interpretability: Revisiting Concept Activation Vectors](https://mlsanity.com/paper/2104.02768/) (2021)
- [Achieving Tractable Minimax Optimal Regret in Average Reward MDPs](https://mlsanity.com/paper/2406.01234/) (2024)
- [Deceptive Alignment Monitoring](https://mlsanity.com/paper/2307.10569/) (2023)
- [Technical Note: Defining and Quantifying AND-OR Interactions for Faithful and Concise Explanation of DNNs](https://mlsanity.com/paper/2304.13312/) (2023)
- [Biased Models Have Biased Explanations](https://mlsanity.com/paper/2012.10986/) (2020)
- ["Why Should You Trust My Explanation?" Understanding Uncertainty in LIME Explanations](https://mlsanity.com/paper/1904.12991/) (2019)
- [Unlocking the Power of LLM Uncertainty for Active In-Context Example Selection](https://mlsanity.com/paper/2408.09172/) (2024)
- [fairret: a Framework for Differentiable Fairness Regularization Terms](https://mlsanity.com/paper/2310.17256/) (2023)
- [What will it take to generate fairness-preserving explanations?](https://mlsanity.com/paper/2106.13346/) (2021)
- [Mitigating Degree Biases in Message Passing Mechanism by Utilizing Community Structures](https://mlsanity.com/paper/2312.16788/) (2023)
- [Exposing the Illusion of Fairness: Auditing Vulnerabilities to Distributional Manipulation Attacks](https://mlsanity.com/paper/2507.20708/) (2025)
- [Efficient Fairness-Performance Pareto Front Computation](https://mlsanity.com/paper/2409.17643/) (2024)
- [Fair Recourse for All: Ensuring Individual and Group Fairness in Counterfactual Explanations](https://mlsanity.com/paper/2601.20449/) (2026)
- [Procedural Fairness and Its Relationship with Distributive Fairness in Machine Learning](https://mlsanity.com/paper/2501.06753/) (2025)
- [Procedural Fairness via Group Counterfactual Explanation](https://mlsanity.com/paper/2603.11140/) (2026)
- [MESD: A Risk-Sensitive Metric for Explanation Fairness Across Intersectional Subgroups](https://mlsanity.com/paper/2603.13452/) (2026)

## Built on by (most cited first, 1 of 1)

- [Transcribe, Then Reason: Two-Pass Decomposition for Multimodal Review](https://mlsanity.com/paper/2609.18958/) (2026)

## Abstract (arXiv)

Machine learning algorithms are being used in high-stakes decisions, including those in criminal justice, healthcare, credit, and employment. The research community has responded with two largely independent research fields: algorithmic fairness, which targets equitable outcomes, and explainable AI (XAI), which targets interpretable reasoning. This survey identifies and maps a novel blind spot at their intersection, which is a model that can satisfy every standard fairness criterion in its outputs while being profoundly unfair in its reasoning process. We refer to this as the procedural bias, and mitigating it requires treating the fairness of explanations as a distinct object of scientific study. To our knowledge, we provide the first unified theoretical and literature review of this emerging field and elucidate the drawbacks of post-hoc explainers in certifying explanation fairness. Our central contribution is a conditional invariance framework formalizing explanation fairness as the requirement that explanations should be indifferent regardless of the protected attributes  P(E(X) ∈ · mid Xrel = xrel,, A = a) = P(E(X) ∈ · mid Xrel = xrel,, A = b) for all task-relevant x, a single principle from which all existing explanation fairness metrics emerge as partial operationalizations. We introduce a seven-dimensional taxonomy, identify three generative mechanisms of explanation inequity (representation-driven, explanation-model mismatch, actionability-driven), and propose a canonical six-step evaluation workflow for operationalizing explanation fairness audits in practice.