AI Promises Concierge Medicine. It Also Evades Federal Review. Call That A Problem.

AI can sift through years of patient history and surface risks that a rushed physician simply cannot catch in a fifteen-minute visit. The catch is that most health AI tools dodge FDA oversight entirely by marketing themselves as wellness products. Mr. Rosenblatt, founder of the longevity company Being and a member of a federal advisory panel convened by CMS, proposes a regulatory crash test. A rotating set of confidential real clinical cases would be presented to any AI Americans consult for medical advice, with results published like car safety ratings.

This illustrates the principle of regulatory arbitrage. Companies exploit gaps between wellness and medical claims to avoid oversight, and the consumer pays the price. The mental model here is the public goods problem. Markets underinvest in safety testing because one firm bears the cost while everyone benefits. Government stepped in for automobiles decades ago with crash tests. Health AI needs the same intervention because information asymmetry between a tech company and a patient is not something the free market corrects on its own.

Mr. Rosenblatt, founder of Being, a health-longevity company, and a member of the HCPLAN Tech-Enabled Health Care Workgroup, a federal advisory panel convened by the Centers for Medicare and Medicaid Services, is advocating for this framework.