It Takes (At Least) Two to Tango in the Rhythm of AI-Enabled Discrimination: How the AI Act Complements EU Non-Discrimination Law?

Despite the elaborate equality and non-discrimination legislation in the European Union (EU), the current legal framework has been widely deemed ill-suited to properly address discriminatory instances that may emerge from the use of algorithms and Artificial Intelligence (AI) technologies. Nevertheless, the potential synergies between the EU Artificial Intelligence Act (AI Act) and non-discrimination law remain underexplored. This article suggests that the AI Act may complement EU non-discrimination rules for the purpose of combatting AI-enabled discrimination in a threefold manner: a) by prohibiting certain AI systems that are prone to produce discriminatory outcomes; b) by regulating the requirements that AI systems need to comply with in order to minimise the risk of discrimination; and c) by enabling the persons affected by discriminatory effects to seek legal protection. Each of these prohibitive, regulatory, and enabling functions of the AI Act are examined in turn, with emphasis placed on their interplay with the existing non-discrimination legislation at EU level. Finally, the article concludes that, apart from the significant complementarities between the two legal regimes both at the level of substantive protection granted to individuals and at the level of enforcement, there are other pieces of EU legislation implicated by the AI Act that may also be applicable when addressing AI-enabled discrimination. Keywords: AI Act, non-discrimination law, EU law, bias, complementarity DOI:10.3935/cyelp.21.2025.615 This work is licensed under the Creative Commons Attribution − Non-Commercial − No Derivatives 4.0 International License. Suggested citation: K Lamprinoudis, ‘It Takes (At Least) Two to Tango in the Rhythm of AI-Enabled Discrimination: How the AI Act Complements EU Non-Discrimination Law?’ (2025) 21 CYELP [ONLINE FIRST].