UK Is Urged to Overhaul Regulation of AI-Medical Devices

The UK needs to change how it regulates AI-enabled medical devices to better protect patients, and should implement a staged approval process that includes continuous monitoring, according to an advisory panel.

AI-driven medical products are a major opportunity for the UK and access needs to be sped up, the National Commission into the Regulation of AI in Healthcare said. Yet rather than a single approval — as is typical with drugs and medical devices — new AI models should be deployed under close supervision, with fuller authorization granted only after they’ve shown real-world safety and efficacy, the commission said in its report.

The findings, which included a consultation with around 12,000 people, are now being considered by the government. The independent commission was established by the Medicines and Healthcare products Regulatory Agency, the UK’s medical watchdog, last September.

Regulators globally are reckoning with how to assess new health products that utilize AI. Devices like AI scribes for doctors, scans that are read by AI, and wearables that monitor patients’ health are increasingly being rolled out. While the devices can lead to greater accuracy and speed up care, questions remain over areas such as safety and potential for bias.

Eventually, AI is likely to rank among the most transformative innovations in healthcare, alongside antibiotics and X-rays, said Alastair Denniston, the chair of the commission and a professor of regulatory science and innovation at the University of Birmingham.

“But as with any powerful technology, we need to understand how to use it safely,” Denniston said. If this isn’t done at pace, the UK will miss out on opportunities, but could also allow technologies to be used before they are ready, he said.

Patients should have easy access to safety information about AI devices, and the MHRA should be given stronger powers to enforce the new rules, the commission said.

The current regulatory system works well for the first-generation of AI devices, but we are now in a new era, said Lawrence Tallon, chief executive officer of the MHRA. “We need to reinvent the regulatory framework for the age of generative AI.”

While the government will respond to the report, Tallon said he expects the MHRA to implement the regulations quickly.