Formalising trade-offs beyond algorithmic fairness: lessons from ethical philosophy and welfare economics

Abstract There is growing concern that decision-making informed by machine learning (ML) algorithms may unfairly discriminate based on personal demographic attributes, such as race and gender. Scholars have responded introducing numerous mathematical definitions of fairness to test the algorithm, many which are in conflict with one another. However, these reductionist representations often bear little resemblance real-life considerations, practice highly contextual. Moreover, metrics tend be implemented within narrow targeted toolkits for algorithm assessments difficult integrate into an algorithm’s broader ethical assessment. In this paper, we derive lessons from philosophy welfare economics they relate contextual factors relevant fairness. particular highlight debate around acceptability inequalities inextricable links between fairness, autonomy. We propose Key Ethics Indicators (KEIs) a way towards providing more holistic understanding whether or not aligned decision-maker’s values.