Abstract
<jats:title>Abstract</jats:title> <jats:p>How can algorithmic systems be effectively governed-especially when they are used in governing? That is, how can groups and individuals whose lives are impacted by algorithmic or automated decision making-including, but not limited to, ‘artificial intelligence’-ensure that those systems and the decisions they produce are valid, appropriate, lawful, and fair? And, in polities and institutional settings aspiring to some degree of democracy, how can affected groups and individuals participate meaningfully in shaping and monitoring the design, operation, evolution, and decommissioning of those systems? While automated decision-making systems present some novel regulatory challenges-for example, in some cases, their complexity or even opacity-already-existing regulatory frameworks, such as anti-discrimination and data protection law, apply to them in many jurisdictions. Additionally, use of algorithmic systems in particular domains, such as employment, justice, government decision making, migration, and warfare is governed by existing domain-specific regulatory frameworks, such as employment and labour law, public law, and international humanitarian and human rights law. The diverse contributors to this Handbook address the question of ‘algorithmic governance’-how to govern algorithmic systems, especially those used in governing-through detailed and closely analysed explorations of specific domains and jurisdictions. The Handbook has three Parts. Part I considers specific regulatory and theoretical lenses: data protection, anti-discrimination, risk management, political economy, rule of law, and the radical legal tradition. Part II considers specific domains, including public law, private law, criminal justice, migration, work, and warfare. Part III offers regional perspectives from Africa, the EU, and Brazil.</jats:p>