ABSTRACT
Artificial intelligence is increasingly integrated into clinical medicine, especially in radiology where it promises more accurate diagnoses and streamlined workflows. However, the integration of AI into healthcare systems raises complex questions about legal liability when AI-assisted decisions result in patient injury. Who should be liable? Legal scholars have proposed a range of tort reforms to address liability concerns, including various forms of strict enterprise liability to hold supply-chain actors liable. These proposals often view tort law as a regulatory tool to deter unwanted conduct and to compensate injured victims. But they also stretch tort principles beyond their doctrinal bounds.
This Article draws on civil recourse theory to contend that modern tort theories proposing tort reform do not serve the purpose of tort law. The purpose of tort law is to empower only injured victims of legally recognized wrongs to hold the tortfeasors accountable. Tort law is about wrongs and redress. Tort liability does not simply arise from harm but from a breach of a relational legal directive owed by the wrongdoer to the victim. This legal directive includes a duty on the wrongdoer not to violate the correlated right of the victim. If a patient cannot show that a physician, or any other ‘responsible’ actor or entity, breached their duty, then no tort has been committed, regardless of the injury. Accordingly, the patient has no claim in torts for compensation. Calls for strict enterprise liability and tort reform fail to remain faithful to the foundational principles of tort law.
While it seems harsh to hold that the patient has no cause of action for her injury, this Article proposes a more pragmatic solution that dispenses with theory to offer victims a straightforward path to compensation: a no-fault compensation program for medical AI injuries modeled after the successful National Vaccine Injury Compensation Program. This approach ensures that patients harmed by medical AI systems receive compensation without having to overcome the evidentiary hurdles of the tort system. By shifting attention away from individual liability and toward no-fault compensation, this Article agrees with legal scholars, medical practitioners, and researchers that the benefits of medical AI are significant. However, actors involved in the development, production, deployment, and use of AI systems should not be deterred by liability concerns or economic disincentives. At the same time, this Article seeks to defend traditional tort law and its conceptual foundations.
Sie, Peter, Medical AI, Civil Recourse Theory, and No-Fault Compensation (October 7, 2025).
Leave a Reply