Assessing Model-Agnostic XAI Methods against EU AI Act Explainability Requirements

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Sovrano, Francesco, Vilone, Giulia, Lognoul, Michael
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918452929757184
author Sovrano, Francesco
Vilone, Giulia
Lognoul, Michael
author_facet Sovrano, Francesco
Vilone, Giulia
Lognoul, Michael
contents Explainable AI (XAI) has evolved in response to expectations and regulations, such as the EU AI Act, which introduces regulatory requirements on AI-powered systems. However, a persistent gap remains between existing XAI methods and society's legal requirements, leaving practitioners without clear guidance on how to approach compliance in the EU market. To bridge this gap, we study model-agnostic XAI methods and relate their interpretability features to the requirements of the AI Act. We then propose a qualitative-to-quantitative scoring framework: qualitative expert assessments of XAI properties are aggregated into a regulation-specific compliance score. This helps practitioners identify when XAI solutions may support legal explanation requirements while highlighting technical issues that require further research and regulatory clarification.
format Preprint
id arxiv_https___arxiv_org_abs_2604_09628
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Assessing Model-Agnostic XAI Methods against EU AI Act Explainability Requirements
Sovrano, Francesco
Vilone, Giulia
Lognoul, Michael
Computers and Society
Artificial Intelligence
Explainable AI (XAI) has evolved in response to expectations and regulations, such as the EU AI Act, which introduces regulatory requirements on AI-powered systems. However, a persistent gap remains between existing XAI methods and society's legal requirements, leaving practitioners without clear guidance on how to approach compliance in the EU market. To bridge this gap, we study model-agnostic XAI methods and relate their interpretability features to the requirements of the AI Act. We then propose a qualitative-to-quantitative scoring framework: qualitative expert assessments of XAI properties are aggregated into a regulation-specific compliance score. This helps practitioners identify when XAI solutions may support legal explanation requirements while highlighting technical issues that require further research and regulatory clarification.
title Assessing Model-Agnostic XAI Methods against EU AI Act Explainability Requirements
topic Computers and Society
Artificial Intelligence
url https://arxiv.org/abs/2604.09628