Gender Bias in AI Systems: A Critical Analysis of Regulatory Frameworks and Policy Responses

Abstract

The rapid proliferation of artificial intelligence systems has exposed pervasive gender biases that reflect and amplify existing societal inequalities, posing significant threats to gender equality and women’s fundamental rights. This article examines gender bias in AI systems through both theoretical and regulatory lenses, analysing how these biases manifest and can be addressed through comprehensive policy frameworks. The first section provides a systematic literature review exploring how bias becomes embedded in algorithmic systems through biased training data, algorithmic design choices, and broader cultural contexts. The second section examines policy responses, comparing UNESCO’s comprehensive recommendations with the European Union’s Artificial Intelligence Act and referencing the Council of Europe Framework Convention on Artificial Intelligence. This analysis reveals a significant disconnect between aspirational frameworks and practical implementation, demonstrating that existing regulatory approaches inadequately address gender bias in AI and highlighting the urgent need for com prehensive integration of gender equality considerations into AI governance frameworks.