Beyond Data: Human Rights, Ethical and Social Impact Assessment in AI (Information Technology and Law Series, 36)

This open access book focuses on the impact of Artificial Intelligence (AI) on individuals and society from a legal perspective, providing a comprehensive risk-based methodological framework to address it. Building on the limitations of data protection in dealing with the challenges of AI, the author proposes an integrated approach to risk assessment that focuses on human rights and encompasses contextual social and ethical values. The core of the analysis concerns the assessment methodology and the role of experts in steering the design of AI products and services by business and public bodies in the direction of human rights and societal values. Taking into account the ongoing debate on AI regulation, the proposed assessment model also bridges the gap between risk-based provisions and their real-world implementation. The central focus of the book on human rights and societal values in AI and the proposed solutions will make it of interest to legal scholars, AI developers and providers, policy makers and regulators. Alessandro Mantelero is Associate Professor of Private Law and Law & Technology in the Department of Management and Production Engineering at the Politecnico di Torino in Turin, Italy. Foreword 7 Preface 12 Contents 17 About the Author 20 1 Beyond Data 21 Abstract 21 1.1 Introduction 22 1.2 Rise and Fall of Individual Sovereignty Over Data Use 23 1.3 Reconsidering Self-determination: Towards a Safe Environment 30 1.4 A Paradigm Shift: The Focus on Risk Assessment 33 1.5 HRESIA: A Multi-layered Process 35 1.6 The Role of Experts 39 1.7 Assessing the Impact of Data-Intensive AI Applications: HRESIA Versus PIA/DPIA, SIA and EtIA 40 1.8 The HRESIA and Collective Dimension of Data Use 47 1.9 Advantages of the Proposed Approach 50 1.10 Summary 50 References 52 2 Human Rights Impact Assessment and AI 64 Abstract 64 2.1 Introduction 65 2.2 A Legal Approach to AI-Related Risks 67 2.3 Human Rights Impact Assessment of AI in the HRESIA Model 70 2.3.1 Planning and Scoping 71 2.3.2 Data Collection and the Risk Analysis Methodology 73 2.4 The Implementation of the Model 79 2.4.1 A Case Study on Consumer Devices Equipped with AI 80 2.4.1.1 Planning and Scoping 80 2.4.1.2 Initial Risk Analysis and Assessment 84 2.4.1.3 Results of the Initial Assessment 88 2.4.1.4 Mitigation Measures and Re-assessment 89 2.4.2 A Large-Scale Case Study: Smart City Government 95 2.5 Summary 102 References 104 3 The Social and Ethical Component in AI Systems Design and Management 111 Abstract 111 3.1 Beyond Human Rights Impact Assessment 112 3.1.1 The Socio-ethical Framework: Uncertainty, Heterogeneity and Context Dependence 114 3.1.2 The Risk of a ‘Transplant’ of Ethical Values 115 3.1.3 Embedding Ethical and Societal Values 119 3.1.4 The Role of the Committee of Experts: Corporate Case Studies 122 3.2 Existing Models in Medical Ethics and Research Committees 128 3.2.1 Clinical Ethics Committees 128 3.2.2 Research Ethics Committees 130 3.2.3 Ethics Committees for Clinical Trials 135 3.2.4 Main Inputs in Addressing Ethical and Societal Issues in AI 137 3.3 Ad Hoc HRESIA Committees: Role, Nature, and Composition 139 3.4 Rights-Holder Participation and Stakeholder Engagement 145 3.5 Summary 148 References 150 4 Regulating AI 156 Abstract 156 4.1 Regulating AI: Three Different Approaches to Regulation 157 4.2 The Principles-Based Approach 159 4.2.1 Key Principles from Personal Data Regulation 161 4.2.1.1 Primacy of the Human Being 163 4.2.1.2 Human Control and Oversight 163 4.2.1.3 Participation and Democratic Oversight on AI Development 164 4.2.1.4 Transparency and Intelligibility 165 4.2.1.5 Precautionary Approach and Risk Management 166 4.2.1.6 Accountability 167 4.2.1.7 Data Minimisation and Data Quality 168 4.2.1.8 Role of Experts and Participation 168 4.2.1.9 Algorithm Vigilance 168 4.2.2 Key Principles from Biomedicine Regulation 169 4.2.2.1 Primacy of the Human Being 170 4.2.2.2 Equitable Access to Health Care 171 4.2.2.3 Acceptability 171 4.2.2.4 Principle of Beneficence 171 4.2.2.5 Private Life and Right to Information 172 4.2.2.6 Professional Standards 172 4.2.2.7 Non-discrimination 173 4.2.2.8 Role of Experts 173 4.2.2.9 Public Debate 173 4.2.3 A Contribution to a Future Principles-Based Regulation of AI 175 4.3 From Design to Law – The European Approaches and the Regulatory Paradox 176 4.3.1 The Council of Europe’s Risk-Based Approach Centred on Human Rights, Democracy and Rule of Law 178 4.3.2 The European Commission’s Proposal (AIA) and Its Conformity-Oriented Approach 183 4.4 The HRESIA Model’s Contribution to the Different Approaches 191 4.5 Summary 193 References 194 5 Open Issues and Conclusions 201 Abstract 201 5.1 Addressing the Challenges of AI 202 5.2 The Global Dimension of AI 204 5.3 Future Scenarios 207 References 211 Index 214 Front Matter Beyond Data Human Rights Impact Assessment and AI The Social and Ethical Component in AI Systems Design and Management Regulating AI Open Issues and Conclusions Back Matter