How to use the Artificial Intelligence Act to investigate AI bias and discrimination: A guide for Equality Bodies

Artificial Intelligence (AI) is transforming our societies, but it also brings risks of bias and discrimination. The Artificial Intelligence Act (AIA)—the first international legal framework dedicated to AI regulation—introduces new safeguards to protect fundamental rights. Equinet’s latest publication, authored by Brent Mittelstadt and financed by the German Federal Anti-Discrimination Agency (FADA), serves as a practical guide for Equality Bodies, equipping them with the tools and knowledge needed to effectively monitor and challenge discriminatory AI systems.

Key Highlights of the Guide:

• Understanding AI Bias: it explains how AI systems can perpetuate discrimination and what fairness metrics exist to assess them.
• Equality Bodies’ Role: it details how Equality Bodies can use new powers under the AIA, including access to technical documentation and testing rights.
• Investigating High-Risk AI: it provides guidance on how to identify, document, and report algorithmic discrimination.
• Collaboration with Regulators: it emphasizes the need for close cooperation with Market Surveillance Authorities (MSAs) to hold AI providers accountable.
• Best Practices & Challenges: it outlines strategies to counter ethics-washing, ensure fairness measures align with legal standards, and address data gaps.

This guide is an essential resource for National Equality Bodies, policymakers, and legal practitioners working to ensure AI systems uphold non-discrimination and fundamental rights.

Download the accessible version of “How to use the Artificial Intelligence Act to investigate AI bias and discrimination: A guide for Equality Bodies” here.