The applications of artificial intelligence (AI) in health care are vast and continue to grow, from automating scheduling and documentation to providing patient education and beyond. One area of growing interest is using AI-enabled chatbots to support care-delivery across a variety of health care contexts. An example of this new patient-facing technology is being developed by Hippocratic AI; the startup has raised $141 million in funding recently to support their efforts to use AI to address health care staffing limitations.
In collaboration with health care providers and health systems, Hippocratic AI has developed AI agents that interact directly with patients to assist with a variety of clinical tasks such as wellness coaching, administration of questionnaires, activity planning, pre- and postoperative follow-up, and much more. Many of the existing AI-agents support physical health care, but offerings have recently expanded to include postpartum mental health. Learn more about Hippocratic AI’s work.
AI continues to evolve in its sophistication and potential uses, and in the absence of comprehensive federal legislation, there continues to be much discussion about how best to ensure the safe use of AI in health care. The U.S. Food and Drug Administration (FDA) has oversight over AI-enabled software and devices that make medical claims, such as the ability to diagnose, treat, or prevent conditions. However, there are many other products, like documentation, scheduling, and education support tools, that fall outside of FDA’s purview.
Researchers recently proposed using a framework that is modeled after the way that clinical laboratories are regulated. This involves federal oversight through Centers for Medicare and Medicaid Services and collaboration with local experts. Safety is ensured through a variety of requirements including certification and set standards for the facilities, personnel, testing procedures, and required documentation. This framework envisions the creation of independent, third-party organizations that would evaluate the AI tools locally, at a specific health care site, to ensure they are used in safe, effective, and equitable ways.
While there is not yet agreement on the best way to approach regulating AI, this will continue to be an important subject as AI’s impact continues to grow. Learn more about the proposed laboratory model.
Find more information about the basics of AI in mental health care.
