Facial Recognition Technology and Potential for Bias and Discrimination
Marcus Smith, Monique Mann · Cambridge University Press eBooks · 2024
This chapter examines facial recognition technology (FRT) and its potential for bias and discrimination against racial minorities in the criminal justice system. The chapter argues that by defining the technology as an automated process, there is an implied objectivity, suggesting that such technologies are free from errors and prejudices. However, facial recognition is dependent on data used to train an algorithm, and operators make judgements about the wider social system and structures it is deployed within. The algorithms that underpin FRT will continue to advance the status quo with respect to power relations in the criminal justice system, unless both data-based and societal-based issues of inequality and discrimination are remedied. FRT is imbued with biases that can negatively impact outcomes for minority groups. The chapter argues that there is a need to focus on systemic discrimination and inequality (rather than calling for a ban of the technology). While the data-based issues are more straightforward to address, this alone will not be sufficient: addressing broader and more complex social factors must be a key focus in working towards a more equal society.