Do We Expect Bias and Discrimination in Algorithmic Decision-Making?

The topic of unfair bias and discrimination fosters intense discussions about possible societal risk associated with implementing algorithmic decision-making systems in high-stakes application contexts (eg, hiring, credit scoring, legal decision-making). In this presentation, I will highlight perspectives on possible unfair bias and discrimination in management contexts of four of the main stakeholders in algorithmic decision-making: providers, regulators, people affected by algorithmic decisions, and users. For the latter two stakeholders groups, I will present results from stakeholder studies showing how humans perceive algorithmic decision-making and possible unfair bias in system outputs.