The use of artificial intelligence in medicine is accelerating rapidly, and large companies — from Google, to Amazon, to Microsoft — are dedicating huge sums to developing novel products that offer big rewards, and equally large risks.
The Food and Drug Administration has already approved three products this year that use AI to help diagnose health problems, including one Thursday that detects wrist fractures. And some companies, like IBM, have put their products on the market without agency signoff.
The technology is evolving so quickly that experts fear the FDA, which is still trying to develop a process for regulating artificial intelligence, seven years after it began the effort in some product categories, won’t be able to keep up. They say the agency can’t compete with private-sector salaries to attract people with the necessary AI knowledge to evaluate the benefits and dangers these algorithms pose.