Algorithmic Discrimination: Continuation of Human Bias or a Gateway to Equality?

In this paper, the question is posed whether Artificial Intelligence (AI) represent an extension of human biases or a path towards achieving equality. It systematically examines three pivotal aspects of algorithmic discrimination: the propensity of algorithms to inherit biases from pre-existing data, the ability of algorithm creators to mitigate this bias through internal rules, and the role of supranational management and regulation in curbing algorithmic bias. By exploring examples of biased models, the research identified instances of both good and bad practices. Additionally, it pointed towards the potential framework for ensuring an equal gateway, as exemplified in the latest international legal frameworks addressing AI non-discrimination principles. The focus of this work is a discussion on the legal regulation of decision-making by algorithms, aiming to prevent potential discrimination.