The EU AI Act in Development Practice: A Pro-justice Approach

2025-04-24

AI Summary

Compliance with the EU AI Act faces operational challenges in risk assessment and fundamental rights protection. Method: This study introduces the High-risk AI Act Toolkit (HEAT), grounded in feminist ethics, which systematically integrates feminist critiques of epistemic authority, stakeholder empowerment, and environmental justice into AI regulatory compliance frameworks. It proposes a “non-neutrality-oriented” methodology to explicitly address structural power asymmetries and value trade-offs in technology governance. HEAT combines normative analysis, participatory design, ethical mapping, and operationalizable checklists to contextualize regulatory requirements. Contribution/Results: Empirical validation demonstrates that HEAT significantly deepens compliance rigor, broadens the scope of fundamental rights impact assessment, and enhances cross-role collaborative efficacy—thereby advancing AI regulation from procedural conformity toward substantive justice.

Technology Category

• Philosophy and Ethics of AI: Bias, Fairness & Equity
• Humans and AI: AI for Accessibility
• Machine Learning: Ethics, Bias, and Fairness

Application Category

• Economics, Online Markets and Human Computation: Fairness and ethical considerations in crowd work and in human-in-the-loop AI systems
• Responsible Web: Fairness, accountability, transparency and ethics of web technologies
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

With the adoption of the EU AI Act, companies must understand and implement its compliance requirements -- an often complex task, especially in areas like risk management and fundamental rights assessments. This paper introduces our High-risk EU AI Act Toolkit(HEAT), which offers a pro-justice, feminist ethics-informed approach to support meaningful compliance. HEAT not only helps teams meet regulatory standards but also translates ethical theory into practice. We show how feminist perspectives on expertise, stakeholder engagement, accessibility, and environmental justice inform HEAT's methods and expand on the Act's baseline. The theories we draw on are not naively utopian. Instead, they inspire non-innocent approaches to interrogating normativity in systems of all kinds -- technological and otherwise. By this we mean that pro-justice orientations are cognizant of their involvement in the systems they seek to critique, and offer best practices with how to grapple with the trade-offs inherent in reducing and eradicating harmful behaviour from within. These best practices, as we explain in this paper, are what HEAT both embodies and enables.

Problem

• Addressing EU AI Act compliance complexity in risk management
• Translating feminist ethics into practical regulatory standards
• Reducing harmful behavior via pro-justice system critiques

Innovation

• HEAT toolkit for EU AI Act compliance
• Feminist ethics-informed pro-justice approach
• Non-innocent normativity interrogation methods