The questions inform the answers: Predictive analytics, risk and legal inequalities
Presented by: Professor Kelly Hannah-Moffat (University of Toronto)
There are multiple forms of algorithmic governance produced through the assemblage of legal logics and risk-informed technologies. These resultant forms governance both reflect and (re)structure social, legal, and criminal justice spheres. The emergence of Artificial intelligence (AI) and, more specifically, machine learning analytics fuelled by big data, is poised to alter some criminal justice practices. The potential of these technologies and the forms of algorithmic governance that are produced have simultaneously enthused and alarmed scholars, advocates, and practitioners. Harnessing the ability of AI creates new possibilities, but it also risks reproducing the status quo and further entrenching existing inequalities. Professor Hannah-Moffat argues that the questions asked, the data sources used, and the organizational logics within which these technologies are deployed fundamentally limit the potential contributions of AI within the criminal justice system. Using examples from predictive policing, AI-informed risk algorithms and AI-enhanced legal search analytics, Professor Hannah-Moffat examines the claims that AI can improve oversight, increase the ‘fairness’ of decisions, reduce bias, and optimize decision-making. Finally, Professor Hannah-Moffat shows how new technologies can make inequalities and ‘data harms’ transparent, while nonetheless still black-boxing and reproducing them in solutions.
Kelly Hannah-Moffat is a Professor in Criminology & Sociolegal Studies and Vice President of People Strategy, Equity & Culture at the University of Toronto. Her interdisciplinary scholarship focuses on penality, criminal records disclosures, big data algorithms, specialized courts, solitary confinement, and institutional risk management practices, including how those processes produce gender and racial inequalities. Her most recent research examines how big data and machine learning technologies are influencing criminal justice decision-making focuses and how forms of algorithmic risk governance are shifting with the emergence of new knowledge brokers and information activists.
Professor Hannah-Moffat also serves on several international editorial boards and is the former co-editor of Punishment and Society (with Mona Lynch). She was a policy advisor and expert witness for several cases related to the use of segregation, deaths in custody, risk assessment, strip searches and conditions of confinement. She works with government and non-profit organizations on criminal justice reform and prison oversight. In 2019, she was awarded the Law and Society Association International Prize for significant contributions to the advancement of knowledge in the field of law and society.
Some of her recent publications include:
- Struthers Monford, K. and K. Hannah-Moffat (2020). The Veneers of Empiricism: Gender, race, and prison classification. Aggression and Violent Behaviour. [pre-print online]
- Avila, F., P. Maurutto and K. Hannah-Moffat (2020). The Seductiveness of Fairness: Are Algorithms the Answer? In M. Schuilenburg and R. Peeters (eds.) The Algorithmic Society. (pp. 87-103). London: Routledge.
- Hannah-Moffat, K and K. Struthers Montford (2019) ‘Unpacking Sentencing Algorithms Risk, Racial Accountability and Data Harms” in Predictive Sentencing Normative and Empirical Perspectives Edited by Jan W de Keijser, Julian V Roberts & Jesper Ryberg. London: Hart Publishing.
- Hannah-Moffat, K. (2019). Algorithmic risk governance: big data analytics, race and information activism in criminal justice debates. Theoretical Criminology, 23(4): 453-470.