Beyond Mercy: Arguments for Allowing (Some) Rule-breaking to Go Unpunished

• In Event: Rethinking Challenges of Technology: AI, Algorithms, and Biopower

Abstract

Advances in surveillance technology and machine learning have opened up new possibilities for the accurate, cheap, and thorough enforcement of rules. For example, traffic cameras can detect and punish driving violations on a scale that human officers could never achieve, machine learning algorithms can help detect fraudulent financial transactions, and CCTV and facial recognition can make it nearly impossible to hide from justice for long. It is natural to assume that more efficient enforcement of fair rules is a good thing, and often it is–unpunished violations can threaten the stability, fairness, and legitimacy of a system of rules. Even so, I argue that there are underappreciated costs to a more efficient enforcement regime, costs that go beyond worries about biased algorithms or privacy rights. Drawing on the work of neo-republican theorists of freedom as non-domination, as well as Susan Wolf’s critique of “moral saints,” I argue that there are compelling moral reasons to allow some detectable wrongdoing to go unpunished. Freedom, fairness, and the legitimacy of rule-based decision-making itself are potentially threatened by the expansion of rule enforcement capacity.

Author

• Brian Palmiter, Harvard University