ChatGPT is penetrating healthcare with explosive might, a US state is trying human-less prescription refills—and more AI news not to miss
If you still think of large-language AI as a newish thing yet to catch on with healthcare consumers, think again.
ChatGPT alone gets rung up for health questions 40 million times a day. The pride of OpenAI also guides 200 million regular users on health-related issues every week. ChatGPT estimates its weekly active user base at 800 million globally, meaning 1 in every 4 individuals in that wildly diverse sample set turns to AI for help with personal health matters. The figures are from a report just released by OpenAI. Among the company’s other noteworthy disclosures:
- More than 5% of all ChatGPT messages globally are about healthcare, averaging billions of messages each week.
- In the U.S., nearly 2 million messages per week focus on health insurance. Common queries include plan comparisons, pricing insights, claims processes, eligibility requirements and cost-sharing details.
- In underserved rural communities across this country, users send an average of almost 600,000 healthcare-related messages per week.
- Some 7 in 10 healthcare conversations in ChatGPT take place outside of normal clinic hours.
Saying it will release a “full policy blueprint” on AI in healthcare early this year, OpenAI suggests five ways the technology might be used to “solve humanity’s hardest healthcare challenges.” Watch for tech lobbyists aligned with OpenAI’s vision to push Washington for help advancing these measures in 2026.
- Open and securely connect the world’s medical data to speed up scientific discovery.
- Build infrastructure to solve healthcare’s hardest problems and rapidly scale solutions.
- Support workers’ transition into the healthcare professions that will be created and expanded by AI.
- Clarify the regulatory pathway for AI-powered consumer medical devices.
- Clarify the scope of medical device regulation to encourage innovation of AI services that support doctors.
Fleshing out the latter point, OpenAI notes that, currently, the FDA regulates as medical devices certain tools intended for use by physicians in their decision-making.
“Certain of these tools are exempted by the 21st Century Cures Act, but this law predates AI solutions that can benefit providers today, and there are open questions about how AI services can meet the exception,” the report’s authors write. “The FDA should establish new guidance on when AI to support physicians’ independent medical judgment is regulated in today’s environment to promote innovation and responsible use of AI by healthcare providers.”
The report spotlights a handful of healthcare professionals using ChatGPT to great innovative effect. Between the lines it suggests OpenAI is becoming a force for U.S. healthcare to reckon with.
More than a quarter of U.S. health systems, 27%, are ponying up for commercial AI licenses. That’s triple the rate of the overall American economy.
The figure is from a survey recently conducted by Menlo Ventures with Morning Consult. The Wall Street Journal cited the finding in a Jan. 5 article showing that large hospital systems are acting as guinea pigs for all sorts of AI applications in healthcare. Which is to say, at the article does, that the big provider organizations are testing what the technology can do while also “revealing—sometimes via alarming mishaps—where it falls flat.” The latter includes situations in which large language models reference nonexistent research, aka “hallucinated content,” to recommend diagnostic conclusions or therapeutic care paths.
- One leading physician who experienced this wackiness firsthand quickly learned to put AI outputs through the “trust but verify” wringer. “It’s not that I don’t ask ChatGPT medical questions,” Mayo Clinic cardiologist Paul Friedman, MD, tells the WSJ, “but, when I do, I always look for the references, click on them and read the abstracts—at a minimum.”
- And then there’s the real-world worry about physician deskilling. “Any time I outsource my thoughts to something that isn’t my own brain, I’m worried I’m going to lose that muscle memory,” says Memorial Sloan Kettering pathologist Anthony Cardillo, MD.
- Despite these kinds of concerns, health systems say they see tremendous promise—and even necessity—in AI. Hospitals are “looking for ways to deal with persistent worker shortages that can burn out clinicians and delay care, WSJ reporters Te-Ping Chen and Chao Deng write, adding to the literature supporting those observations. “They are also looking for efficiency wherever they can find it while cuts to Medicaid loom.” The reporters quote Doug King, MBA, chief digital and innovation officer at Northwestern Medicine. “When you think about the tsunami of need that’s coming [to our] society,” King says, “technology is one of the only levers we have to pull.”
- Full article here (behind paywall).
One state is experimenting with drug prescriptions ordered by AI without routine oversight by a human.
The program only handles refills. Still, it represents “a high-stakes test of whether AI can safely take on one of healthcare’s most sensitive tasks.” The quote is from Politico reporters Yasmin Khorram and Ruth Reader, whose article looks at a pilot program underway—in Utah, of all places—and at how far stakeholders and regulators might let medical AI go with only light human supervision.
- The test program uses software from the startup Doctronic. So far, the FDA has not weighed in on the development. If the agency decides it would be in the right to regulate this use of AI, the move “could complicate or slow the expansion of” Doctronic-type models, the authors point out. Either way, “By inserting algorithms into one of medicine’s most fundamental relationships, Utah’s initiative could represent the first step in upending how care is delivered in the U.S.”
Also worth your while:
- Does AI belong in the exam room? Lawsuit alleges Sharp violated patient privacy. (San Diego Union-Tribune)
- AI has finger on pulse of [China’s] healthcare advances (China Daily)
- The AI Power List: The most influential people in artificial intelligence (Business Insider)
Dave P. has worked in journalism, marketing and public relations for more than 30 years, frequently concentrating on hospitals, healthcare technology and Catholic communications. He has also specialized in fundraising communications, ghostwriting for CEOs of local, national and global charities, nonprofits and foundations.