Privacy Risks and Preservation Methods in Explainable Artificial Intelligence: A Scoping Review

2025-05-05

AI Summary

This paper addresses the privacy risks arising from explanation dissemination in eXplainable Artificial Intelligence (XAI). Conducting a systematic scoping review of 57 publications (2019–2024) following the Arksey & O’Malley framework, it uncovers an inherent tension between interpretability and privacy preservation. Through thematic coding and cross-study meta-synthesis, the work formally defines “privacy-preserving explanations” via four essential characteristics and proposes a unified classification framework encompassing four risk categories and five technical paradigms. It introduces evaluation dimensions that jointly satisfy interpretability requirements and regulatory compliance (e.g., GDPR). The study delivers actionable design guidelines and evidence-based recommendations for both regulators and practitioners. Its core innovation lies in embedding privacy considerations throughout the entire XAI explanation generation pipeline—enabling a closed-loop transition from risk identification to methodological implementation.

Technology Category

• Humans and AI: Explainable AI (XAI) for Human Understanding
• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability
• Machine Learning: Privacy

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Security and Privacy: Data transparency and provenance
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Explainable Artificial Intelligence (XAI) has emerged as a pillar of Trustworthy AI and aims to bring transparency in complex models that are opaque by nature. Despite the benefits of incorporating explanations in models, an urgent need is found in addressing the privacy concerns of providing this additional information to end users. In this article, we conduct a scoping review of existing literature to elicit details on the conflict between privacy and explainability. Using the standard methodology for scoping review, we extracted 57 articles from 1,943 studies published from January 2019 to December 2024. The review addresses 3 research questions to present readers with more understanding of the topic: (1) what are the privacy risks of releasing explanations in AI systems? (2) what current methods have researchers employed to achieve privacy preservation in XAI systems? (3) what constitutes a privacy preserving explanation? Based on the knowledge synthesized from the selected studies, we categorize the privacy risks and preservation methods in XAI and propose the characteristics of privacy preserving explanations to aid researchers and practitioners in understanding the requirements of XAI that is privacy compliant. Lastly, we identify the challenges in balancing privacy with other system desiderata and provide recommendations for achieving privacy preserving XAI. We expect that this review will shed light on the complex relationship of privacy and explainability, both being the fundamental principles of Trustworthy AI.

Problem

• Identifying privacy risks in AI explanation disclosures
• Reviewing methods for privacy preservation in XAI systems
• Defining characteristics of privacy-compliant AI explanations

Innovation

• Scoping review of privacy risks in XAI
• Categorizes privacy preservation methods
• Proposes characteristics for privacy explanations