The Risks of Using AI in Health Care

Despite enthusiasm about AI’s potential impact on health care, there are downsides that need to be addressed. MIT postdoctoral associate Tom Hartvigsen will join MIT Horizon to discuss two of the issues for use of AI in health care: bias and explainability. Bias can occur when data is not gained from a representative sample of the population, which particularly disadvantages members of minority populations. Explainability deals with the lack of understanding of how and why AI comes to the conclusions it does. After the talk, Hartvigsen will take questions from the audience.