Trust, but Verify: Informational Challenges Surrounding AI-Enabled Clinical Decision Software

September 18, 2020Duke Law News

White Paper

The Center for Innovation Policy at Duke Law (CIP) and the Duke-Margolis Center for Health Policy (Duke-Margolis) have published a new white paper on the use of AI-enabled software in health care. The paper, “Trust, but Verify: Informational Challenges Surrounding AI-Enabled Clinical Decision Software,” funded by a grant from the Greenwall Foundation, focuses on the challenges of using artifcial intelligence (AI) in health care and how to assist software developers, policymakers, regulators, clinical workers, and others to encourage innovation in safe and effective AI-enabled medical products.

The white paper is co-authored by Christina Silcox, managing associatie, Duke-Margolis; Arti K. Rai, Elvin R. Latty Professor of Law and Faculty Co-Director, CIP, Duke Law School; and Isha Sharma, senior research assistant, Duke-Margolis.

Covering a broad range of issues, the report addresses the challenges of using AI in health care, forcusing on AI-enabled clinical decision software and the role of patents and regulation in the adoption of such software. Particular focus is on the benefits and risks of disclosing information during the development and use of such products.

The white paper distills a myriad of research including meetings hosted by CIP and Duke-Margolis and interviews with various stakeholders. The report is released in conjunction with a a forthcoming article in the Journal of Law and the Biosciences (vol. 7, iss. 2), “Accountability, Secrecy, and Innovation in AI-Enabled Clinical Decision Software,” that provides details of the research methodology and additional findings related to disclosures in patents, peer-reviewed publications, and U.S. Food and Drug Administration filings.

The opinions expressed in these papers are those of the authors and do not represent the views or policies of any other organization external to Duke.