As the use of artificial intelligence and machine learning technologies in the medical and health care sector increases, medical professionals, health care facilities and malpractice insurers face new risks, experts say.
While artificial intelligence may help raise the standard of care that patients receive from medical professionals, its involvement in medical decision-making will also give patients new reasons to sue, expanding the potential liabilities confronting doctors, they say.
AI is beginning to play a larger role in the practice of medicine, and as it does the goal of AI is to improve medical decision-making, which will have an effect on medical outcomes, said Dr. Richard Anderson, chairman and CEO of The Doctors Co. in Napa, California.
“To the extent that medical outcomes are better, then everyone’s happy. But to the extent that medical outcomes will occasionally be worse, or if there’s a catastrophic problem with a particular algorithm that could affect many people at once, that creates a nightmare litigation scenario,” Dr. Anderson said.
Medical malpractice insurers are taking note, and as AI-driven technologies become more widespread and lead to claims, insurance coverage will have to evolve, just as it did with the advent of electronic health records and telemedicine, they say.
“With any kind of new technology or new evolution in the provision of care, the question is always going to be the interaction between humans and the technology,” said Greg Ferrell, Austin, Texas-based vice president of professional liability with Admiral Insurance Group, a subsidiary of W.R. Berkley Co.
“How much reliance is there on it? How much of a human factor is there? Software is great and AI is great, but in reality those calculations are only as good as the data input by a human hand. So there is always that concern,” said Mr. Ferrell.
The increased utilization of information technology in health care delivery is blurring the distinctions between providers and patients, said Dale Van Demark, a partner with McDermott, Will & Emery LLP in Washington.
“As individual consumers, we are increasingly being empowered to take care of ourselves with these tools” via digital products and apps that can be put on smartphones, said Mr. Van Demark.
“A good example would be a diabetes management tool … where a consumer, acting alone or in conjunction with a physician, is able to start taking over more responsibility for the management of that chronic condition,” he said.
“Who’s the provider of the health care service? Is it the patient, the doctor or the device? Is it some combination of those?” he said.
In the near term, AI is going to present many legal quandaries, said Scott M. Kosnoff, partner at Faegre, Baker Daniels LLP in Indianapolis.
“If I were to go to the doctor and for whatever reason he does not avail himself of AI and makes a mistake, I would argue that the doctor failed to rely on AI. We may get to the point where a failure to rely on AI is an indicator of a doctor’s — if not malpractice — of not living up to the prevailing standard,” said Mr. Kosnoff.
“You could also argue the other side of that coin,” to the extent that the doctor relied on AI but didn’t exercise his/her own independent medical judgment relying “exclusively on the AI without asking if it was working like it should,” Mr. Kosnoff said.
Going forward, “it’s likely that (patients) will seek liability from providers for injuries that result from the use of AI,” said Nicholson Price, a professor of law at the University of Michigan in Ann Arbor, adding: “I suspect they’ll also sue the makers of AI.”
“A good analogy is autonomous vehicles. Who’s going to be liable for automobile accidents between autonomous vehicles? The same kinds of issues we’re facing in the not-too-distant future with autonomous vehicles are like what we face with AI in medicine,” said Dr. Anderson.
From the underwriting standpoint, the challenge is “how to connect the dots between the technology itself and the true front-line provision of care,” said Mr. Ferrell.
When insurers are unsure about how new technology is involved, they are unable to structure insurance policies that address all the relevant exposures facing providers, he said.
“At the early onset, it’s going to be a learning curve. It won’t be an intentional omission of coverage … but until you know how those suits are going to be brought — the vernacular, the vocabulary, the terminology used, the extent to which technology is built into the process of care — there’s really no way we can say this is how we’ll cover it,” Mr. Ferrell said.
Coverage will change and evolve, he added, and what might eventually follow is a package policy that includes medical malpractice, general liability, products liability, the technology exposure and a cyber component as well, he said.
“My guess is AI will be covered as part of your standard policy as opposed to something distinctly different … As AI becomes more integral in medical practice, it’ll become difficult to insure medical practice except where it’s influenced by AI,” said Dr. Anderson.
