AI in healthcare – why standards, commonsense, and education are essential

In my previous report on AI in healthcare, I revealed how one UK regulator is taking a “guidance heavy, legislation light” approach to the technology – a nimble and pragmatic solution, perhaps, but also one that leaves grey areas for unethical organizations to exploit. A criticism that could apply to any voluntary code, of course, such as the US’ AI Bill of Rights.

So, in such an environment, what barriers exist to ensuring patient safety and evaluating the effectiveness of AI in healthcare? How can organizations balance the benefits against the risks? And what are the options for post-market surveillance – in other words, for monitoring the safety and effectiveness of an AI-powered medical device or service once it has been deployed?

These are important questions, because healthcare is far from the only sector to be sitting on its hands while the whirlwind of hype rages around it. Earlier this month, IT ...