President Joe Biden’s executive order on artificial intelligence is a warning to those developing AI systems in health care: The government's laissez-faire approach is ending. The order — the full text was released last night — envisions, within months, a new regulatory structure that will aim to ensure AI tools already in use across a broad swath of the health care industry maintain “appropriate levels of quality.” Those areas include AI to help doctors evaluate patients and diagnose diseases, but also tools that boost research and development of drugs and medical devices, measure the quality of care, streamline insurance benefits administration, manage the patient experience, and prioritize public health needs. Because AI systems are constantly learning as they process new data, Biden is asking the Department of Health and Human Services to stand up a regulatory structure capable of assessing the tools before they go to market and overseeing their performance after they do. That structure would take whistleblower complaints about safety, privacy and security and catalog errors made by AI systems. Biden wants HHS to ensure that that information is well-publicized to both providers and patients. Spurring innovation: In addition to mitigating the risks of unregulated AI, the order seeks to explore whether AI tools can make care better, cheaper and more equitable. The order tasks HHS with ramping up grantmaking aimed at: — developing AI-enabled tools that create personalized immune-response profiles for patients. Researchers envison future therapies tailor-made to help each patient’s unique immune system fight disease. — improving health care data quality on which AI systems rely. — promoting workplace efficiency and job satisfaction by reducing administrative burdens. — bolstering care for veterans and supporting small businesses through two three-month nationwide “AI Tech Sprint” competitions. What’s next? Biden has given HHS and other agencies six months to a year to carry out the order.
| 
