In this report, the authors describe the basic features of U.S. tort law and analyze their significance for the liability of artificial intelligence (AI) developers whose models inflict, or are used to inflict, large-scale harm. The report is intended to be useful to AI developers, policymakers, and other nonlegal audiences who wish to understand the liability exposure that AI development may entail and how this exposure might be mitigated.
Authors
- Division
- RAND Global and Emerging Risks
- Pages
- 65
- Published in
- United States
- RAND Identifier
- RR-A3084-1
- RAND Type
- report
- Rights
- RAND Corporation
- Series
- Research Reports
- Source
- https://www.rand.org/pubs/research_reports/RRA3084-1.html
Table of Contents
- U.S. Tort Liability for Large-Scale Artificial Intelligence Damages 1
- About This Report 3
- Technology and Security Policy Center 3
- Funding 3
- Author Affiliations 3
- Acknowledgments 3
- Summary 5
- Contents 10
- Introduction 13
- Overview 13
- Methodology 15
- Organization of This Report 15
- Technical Review 16
- The Nature of Tort Law 19
- Tort Law Is Common Law 19
- Tort Law Varies Considerably Across Jurisdictions Including Those Within the United States 20
- Remedies for Torts Compensatory Damages Punitive Damages and Injunctions 23
- Negligence 26
- The Structure of Negligence Law and Its Application to AI Development 26
- Duty 26
- Breach Factual causation Proximate cause 27
- Intervening Agency and Liability for Third-Party Misuse 32
- Causation Issues 36
- The Dynamics of Jury Decisionmaking and Settlement 37
- Conclusion 37
- Products Liability 38
- Are AI Systems Products 39
- Design Defect Liability The Risk-Utility Test 40
- Design Defect Liability The Consumer Expectations Test and the Malfunction Doctrine 42
- Liability for Defective Warnings 44
- Conclusion 45
- Strict Liability for Abnormally Dangerous Activities 47
- Public Nuisance 50
- The First Amendment and Section 230 52
- First Amendment Considerations 52
- The Application of Section 230 to AI 54
- Conclusion 56
- References 58
