What is the problem
With increasing integration of artificial intelligence and machine learning in medicine, there are concerns that algorithm inaccuracy could lead to patient injury and medical liability. Current liability frameworks are inadequate to encourage both safe clinical implementation and disruptive innovation of artificial intelligence. How should policymakers alter the legal liability framework for medical AI?
What are we doing
In Milbank Quarterly (Current liability frameworks are inadequate to encourage both safe clinical implementation and disruptive innovation of artificial intelligence), we proposed one of the first frameworks for liability reform for medical AI. We argued for several potential reforms, including insurance, indemnification, and special/no-fault adjudication systems. In subsequent work in the Harvard Business Review and Scientific American, we have expanded our recommendations for broader classes of AI, including autonomous vehicles.
Our article has been cited in several policy documents about ethical AI.