Applying Existing Legal Frameworks to Predictive Policing with Artificial Intelligence: Gaps, Tensions, and Individual Harms

Artificial intelligence has greatly changed the nature of policing and criminal law enforcement more generally. Predictive policing, though not a new tactic in many jurisdictions, has been quickly advanced with the infusion of artificial intelligence. This has led to problematic applications of fundamental rights. This thesis explores several of these failings: bias and discrimination; loss of the presumption of innocence; and adaptation of the reasonable suspicion standard. The thesis looks into three jurisdictions (the United States, the United Kingdom, and the Council of Europe states) to demonstrate that though these systems vary greatly in their criminal laws and fundamental rights frameworks, the use of predictive policing aided by artificial intelligence produces the same flawed results when applied to policing of individuals and groups.