ABSTRACT
Nowadays, companies feel there’s a business imperative to use AI systems and tools in many of their processes. This includes, for example, recruitment, talent management, decision-support systems, data analysis, predictive analytics, or customer interaction (Black & van Esch, 2020, 2021; Desouza et al., 2020). AI systems are thought to be key to both reducing running costs and also helping companies gain efficiencies and value (Desouza et al., 2020; Forman et al., 2020; Reynolds, 2021; Seiler, 2021), but they come with implementation challenges, especially for those companies that want to ensure that their systems are fair and not perpetuating gender and racial discrimination. This has been the focus of much academic research, which has emphasized the ability of AI to misgender (Keyes, 2018), to oppress (Browne, 2015; Noble, 2018; Woods, 2018), to exclude (Buolamwini & Gebru, 2018), and to stereotype (Kay et al., 2015).
