Research Handbook on Meaningful Human Control of Artificial Intelligence Systems

Edited by Giulio Mecacci, Assistant Professor of Ethics and Philosophy of AI, Donders Institute for Brain, Cognition and Behaviour, Radboud University, the Netherlands, Daniele Amoroso, Professor of International Law, University of Cagliari, Italy, Luciano Cavalcante Siebert, Assistant Professor of Responsible AI, Faculty of Electrical Engineering, Mathematics and Computer Science, Delft University of Technology, David Abbink, Professor of Human-Robot Interaction, the Faculty of Mechanical Engineering and the Faculty of Industrial Design Engineering, Delft University of Technology, Jeroen van den Hoven, Professor of Ethics of Technology, Delft University of Technology and Filippo Santoni de Sio, Professor of Ethics of Technology, TU Eindhoven, the Netherlands

‘The Research Handbook on Meaningful Human Control of Artificial Intelligence Systems provides an excellent collection of scholars from a number of fields who each, in turn, give substance and clarity to the thorny and dynamic concept of Meaningful Human Control (MHC). The Handbook carefully considers the multi-dimensionality of AI and its relationship to human agency, and offers one of the most comprehensive and thoughtful engagement with the ways in which MHC challenges us, while also offering ideas about how these challenges could be addressed, if not overcome. An important and much needed contribution to an ongoing debate.’

• ContentsList of contributors viiAcknowledgements x1 Introduction to meaningful human control of artificially intelligent systems 1Authors (listed in alphabetical order): David Abbink, Daniele Amoroso,Luciano Cavalcante Siebert, Jeroen van den Hoven, Giulio Mecacci andFilippo Santoni de SioPART I PHILOSOPHY AND ETHICS2 Automated mobility and meaningful human control: when and why iscontrol important and what are its different dimensions? 13Sven Nyholm3 Reflection machines and the proximity scale of reasons: addressingaccountability asymmetry 28Pim Haselager and Giulio Mecacci4 Out of control: flourishing with carebots through embodied design 38Anco Peeters5 Meaningful human command: advance control directives as a method toenable moral and legal responsibility for autonomous weapons systems 53S. Kate DevittPART II LAW AND GOVERNANCE6 Legal and governance perspectives on meaningful human control: thecase of automated mobility 82Giuseppe Contissa7 From “human control” in international law to “human oversight” in thenew EU act on artificial intelligence 105Juliane Beck and Thomas Burri8 Meaningful human control in shared medical decision making 133Susanne Beck, Simon Gerndt, David Samhammer and Peter Dabrock9 Meaningful human control: an international criminal law account 150Marta BoPART III DESIGN AND ENGINEERING10 Designing automated vehicle and traffic systems towards meaningfulhuman control 165Simeon Calvert, Stig Johnsen and Ashwin George11 Reflective hybrid intelligence for meaningful human control indecision-support systems 191Catholijn M. Jonker, Luciano Cavalcante Siebert and Pradeep K.Murukannaiah12 Meaningful human control of increasingly autonomous robots for surgery 208Fanny Ficuciello, Mohammad Hossein Hamedani and Guglielmo Tamburrini13 Designing for meaningful human control in military human-machine teams 235Jurriaan van Diggelen, Karel van den Bosch, Mark Neerincx and Marc SteenPART IV INTERDISCIPLINARY AND SYSTEMIC PERSPECTIVES14 Too much control? Health, sex and war 257Ezio Di Nucci15 Worldbuilding a planet-positive future enabled by human + AI coevolution 267Ann Pendleton-Jullian16 Holistic bow-tie model of meaningful human control over effectivesystems: towards a dynamic balance of humans and AI-based systemswithin our global society and environment 313Frank Flemisch, Marcel Baltzer, David Abbink, Luciano Cavalcante Siebert,Jurriaan van Diggelen, Nicolas Daniel Herzberger, Mark Draper, MichaelBoardman, Marie-Pierre Pacaux-Lemoine, Joscha Wasser