As AI seems to invade every field of technology and science, there’s much observation of when it doesn’t work. It’s no secret that it’s not perfect. I’ve recently reported that AI-powered ‘fake news detectors’ can’t accurately function with current datasets. But perhaps more dangerous than putting too much trust in insufficient AI technology is the prospect of malicious individuals sabotaging AI that we trust.
In a March 21 article called “Warnings of a Dark Side to A.I. in Health Care,” The New York Times reported that researchers are concerned about “adversarial attacks” on AI systems that use medical images like MRIs to make diagnoses and evaluations.
“By changing a few pixels on a lung scan, for instance, someone could fool an A.I. system into seeing an illness that is not really there, or not seeing one that is,” Cade Metz and Craig S. Smith wrote in the article.
“The concern is less that hackers might cause patients to be misdiagnosed, although that potential exists. More likely is that doctors, hospitals and other organizations could manipulate the A.I. in billing or insurance software in an effort to maximize the money coming their way.”
The article cited Samuel Finlayson, a Harvard researcher and an author of a paper in Science that raised these concerns. He said the concerns are valid because of the giant amount of money that changes hands in the healthcare industry.
This is yet another sign that AI may be moving at too fast a pace for harm to be limited as much as possible. The public visibly recoiled when a self-driving car killed a pedestrian last year, but that was because it was a sci-fi-esque instance of AI killing a human. It was exactly the thing to cause the public to disproportionately fear AI. But this issue, the one raised in this NYTimes article, is far more frightening because of how little fear it elicits in the public.
As debate rages in the country about how healthcare should be structured and who should pay, most of the public has no idea how it works currently. And as AI is allowed to take over certain parts of the process, it opens the door for people to understand even less about how their health and money are being manipulated.