Applications Open for London Programme Testing AI Medical Devices in NHS Settings
The MHRA, NHS England's London region and health-innovation networks have opened expressions of interest for London Region I, a regulatory sandbox for AI-enabled medical devices.

Artificial-intelligence developers and NHS organisations in London have been invited to apply for a new programme that will allow selected AI-enabled medical devices to be tested in clinical settings under regulatory oversight.
The initiative, known as London Region I, is being delivered by the Medicines and Healthcare products Regulatory Agency, NHS England's London region and London's health-innovation networks.
Although the programme was initially announced on 10 June, the government updated the scheme on 30 July to confirm that expressions of interest are now open.
Up to 10 AI medical-device manufacturers will be selected during the initial phase. Participating businesses will work with London-based NHS providers to deploy and evaluate their technologies in real healthcare environments.
The programme has been described as a regulatory sandbox — a controlled structure in which new technologies can be tested while developers, healthcare providers and regulators work together to identify safety, clinical and compliance issues.
Officials say the programme is intended to produce stronger real-world evidence about whether AI medical devices are safe, effective and capable of improving patient outcomes.
Potential uses of AI-enabled devices include supporting medical imaging, detecting disease, monitoring patients, helping clinicians interpret data and identifying people who may require earlier intervention.
However, the introduction of AI into healthcare carries significant risks. Regulators must consider whether systems perform consistently across different patient groups, whether clinical decisions can be explained, how sensitive health information is protected and what happens when an algorithm produces an incorrect recommendation.
Bias is another major concern. An AI system trained using incomplete or unrepresentative data could be less reliable for particular communities, age groups or ethnic populations.
The London programme is designed to allow these issues to be examined before technologies are adopted more widely across the NHS.
Manufacturers will still be expected to meet applicable medical-device standards. Participation in the sandbox does not amount to automatic approval, nor does it remove the requirement for evidence demonstrating safety and effectiveness.
The MHRA said regulation should support useful innovation rather than unnecessarily delay it, while ensuring that patient protection remains central.
For the NHS, the business case rests on whether AI can improve diagnostic speed, expand access to services, reduce administrative pressure and help overstretched clinical teams manage rising demand.
The programme's longer-term significance will depend on whether successful technologies can move from controlled trials into routine use without creating new risks, costs or inequalities.
The opening of applications marks the operational stage of the initiative and gives London-based NHS providers an opportunity to identify clinical problems that selected AI companies may be able to address.

WHO Introduces New Regulatory Path to Speed Access to Safe Medical Devices
A new interim list of trusted medical-device regulators is designed to help countries rely on established assessments and reduce unnecessary duplication.

Australia Takes Telegram to Court Over Alleged Failure to Remove Terrorist Content
The eSafety Commissioner has begun Federal Court proceedings alleging Telegram failed to promptly remove violent extremist material reported by Australian users. Telegram rejects the claims.

Meta's Free Cash Flow Falls 91% as AI Infrastructure Spending Surges
Meta generated $784 million in free cash flow in the second quarter, down from about $8.55 billion a year earlier, as capital spending on AI data centres accelerated.