An interesting perspectives article appeared online in the Harvard Business Review on July 13, 2021: It addressed the barriers that the U.S. liability system may place on the deployment of artificial intelligence (AI) applications in medicine and other industries. The authors addressed the acceptance of AI technologies in the framework of possible injury reparations reform. A principal topic was the standard of care.
The standard of care is the doctrine most relevant to any professional whose alleged negligence could result in professional liability litigation. Did the professional deviate from the standard of care that a reasonable professional would provide in the same or similar circumstances? Such standards are generally described to malpractice tribunals through the testimony of expert witnesses.
I found myself thinking about the standard of care for the use of AI in diagnostic radiology. What is the standard of care for the profession now? Will it evolve?
The authors advocated adopting and changing standards of care as the means to facilitate the embrace of AI. Changing the standard of care, they wrote, would change how professionals are judged in malpractice litigation, and could make the difference between fighting and facilitating AI.
With respect to radiology, it is my view that a major role for the industry's professional societies – notably through new practice parameters adopted by the American College of Radiology – will likely be key to an evolving standard of care for AI's usage in radiologists' practices. But mostly the standard will be set by the scope of the growing number of AI services that the profession actually delivers.
Curtis Langlotz, M.D., Ph.D., Professor of Radiology and Biomedical Informatics and Director of the Center for Artificial Intelligence in Medicine and Imaging at Stanford famously said, "AI will not replace radiologists, but radiologists who use AI will replace radiologists who don’t." We can expect the standard of care to be set by how widely radiologists utilize AI tools to enhance and improve their interpretive services. Will those radiologists who do not fully embrace AI ultimately find themselves outside the standard of care as radiologist experts may come to describe it?
Although we focus on the U.S. liability system, the principles underlying our recommendations can be applied to many countries. Indeed, ignoring liability anywhere could result in both unsafe deployment of AI and hamper innovation.