NHSX has identified the need for a regulatory sandbox to support the development of artificial intelligence for healthcare.
Its chief executive officer Matthew Gould has pointed to it as one of the priorities for how the service should approach the use of AI following a meeting late in January of the regulators and organisations involved.
This was prompted by concerns that the technology could be used in an unsafe manner, or that clinicians and innovators could be held back by worries over the lack of a regulatory framework in which to work.
“Doing AI right means putting a set of rules around it that will make sure it is done safely, in a way that respects patients’ privacy and keeps the confidence of citizens and staff,” Gould said.
“Some of these rules already exist, like the 2018 Data Protection Act, for example, which put the General Data Protection Regulation into UK law. But there are gaps, lots of regulators on the pitch, and a lack of clarity on both standards and losses.”
The meeting between bodies including the Medicines and Healthcare products Regulatory Agency (MHRA), the National Institute for Clinical Excellence (NICE) and the Care Quality Commission (CQC) led to the adoption of priorities that includes the regulatory sandbox.
Safe space
This would bring together all of the sandbox initiatives from different regulators to provide innovators in AI with a safe space to develop and test their systems for healthcare.
It would be accompanied by a capability to assess the systems at the pace and scale required, through either the relevant regulators or designated organisations working to clear standards. Gould emphasised the potential of the MHRA in this area, while saying it should also be responsible for regulating the safety of AI systems.
Along with this NICE would be responsible for assessing their value to determine whether they should be deployed, the HRA for overseeing the research to generate evidence, and the CQC to ensure that providers are following best practice.
Another priority is to work out how to handle machine learning. “We know it’s difficult, but we need to develop a proposal, test it, iterate and keep iterating,” Gould said, adding that NHSX will initially take the lead in the effort.
Others are to set up a platform to bring together all regulatory standards and create a single point of contact, and maintain communications with clinicians, innovators and the public.
Gould also highlighted the role of the NHS AI Lab, set up last August, partly as a source of funding for initiatives.
Guidelines and weaknesses
The move has come three months after NHSX published guidelines on the use of AI in healthcare, and on the week that the Committee on Standards in Public Life highlighted governance weaknesses for the use of AI in public services.
“It’s a huge agenda,” he said. “But it really matters and we need to move this forward at pace.
“The prize – if we can get this right – is making the UK a world leader in AI for health, giving the NHS the benefits of this new technology safely.”
Image from GOV.UK, Open Government Licence v3.0
