Algorithms in the Workplace, an Unbiased Technology?

Abstract

This study analyses how biases may emerge in algorithms due to human intervention and explores the legal risks associated with these biases, conceptualising algorithm as a sufficiently detailed and systematic instruction of action to solve a mathematical problem so that, when implemented correctly, the computer computes the correct output for each correct set of inputs. The findings highlight several specific risks of algorithmic bias without explicit intent. Additionally, the limited case law reviewed indicates a growing but still underdeveloped judicial recognition of these risks. The results underscore a need for legal frameworks to more explicitly address algorithmic transparency and accountability, especially where human biases may disproportionately affect certain demographic groups or individual rights.