AI Breakthroughs Drive Expansion of ‘Airlock' Testing Programme to Support AI-Powered Healthcare Innovation
A £1 million boost to the Medicines and Healthcare products Regulatory Agency's (MHRA) pioneering AI Airlock programme will expand access to a first-of-its-kind regulatory testing ground where companies can work directly with regulators to safely test new AI-powered medical devices and explore how to bring them to patients faster, through streamlined regulations.
Applications for the second round of the programme have opened and follow a successful pilot phase that saw four breakthrough AI technologies, including software that could help doctors create personalised cancer treatment plans, and a tool to help hospitals, AI developers, and regulators monitor AI performance in real time, tested in a regulatory 'sandbox' environment.
Similar to an airlock on a spacecraft, the 'sandbox' testing space creates a boundary between experimental AI and fully approved medical technology used in the real world.
This initiative builds on commitments in the UK Government's AI Opportunities Action Plan and the government response to the Regulatory Horizons Council report on regulation of AI as a medical device to enable safe AI innovation through strategic guidance to regulators and enhance their AI capabilities.
This programme is backed by the UK Government's new Regulatory Innovation Office (RIO), which is supporting regulators to test more agile, flexible ways of working that can keep pace with emerging technologies like AI.
Science Minister, Lord Vallance, said:
'Backing innovation means backing better regulation – and that's what the RIO is here to do.
'Smarter, faster approaches like the AI Airlock are helping to cut red tape, bring safe new technologies to patients quicker, and ease pressure on our NHS – fuelling the Government's Plan for Change.'
Health Minister, Baroness Merron, said:
'AI has huge potential to improve healthcare, and we need to use it safely and responsibly. The AI Airlock programme is a great example of how we can test new technology thoroughly while still moving quickly.
'This £1 million investment will help bring new medical tools to patients faster and strengthen the UK's position as a global leader in healthcare innovation.'
Those selected for the next round of the AI Airlock programme will be able to test their AI healthcare products under careful supervision allowing for regulatory challenges to be identified early and adjustments made.
James Pound, MHRA Interim Executive Director, Innovation and Compliance, said:
'Traditional regulatory pathways weren't designed with AI's unique characteristics in mind – including its capacity to analyse large quantities of data and help automate existing manual processes. The AI Airlock programme helps address this gap by creating a supervised testing ground where these novel technologies and challenge areas can be safely investigated.
'The technologies and devices which have been evaluated to date have shown the limitless potential of AI to improve patient outcomes, free up NHS resources, and enhance the accuracy and efficiency of healthcare services.
'With AI, we must balance robust oversight with flexibility that doesn't stifle innovation, and this programme achieves that balance.'
Four projects were selected for the inaugural AI Airlock cohort, each focused on addressing critical healthcare challenges using AI. Among them was health technology multinational Philips' Radiology Auto Impression project which tested the use of generative AI to automate the writing of radiologists' final impressions – a critical section of radiology reports that summarises key findings from imaging procedures.
Working directly with MHRA experts through weekly meetings, the team gained valuable insights about the need to involve their end users – radiologists – to help define testing strategies. As Yinnon Dolev, Philips' Advanced Development NLP (Natural Language Processing) Tech Lead noted, the collaboration with regulators was 'almost unheard of' and provided 'a catalyst for meaningful progress expediting our development activities.'
OncoFlow, another first round project, looked at the use of AI to help healthcare professionals create personalised management plans for cancer patients, with the potential to reduce waiting times for cancer appointments, leading to earlier treatment and the possibility of significantly increasing patients' chances of survival. Co-founder Aruni Ghose said the Airlock programme provided his team with the chance to validate the product in a simulated clinical setting and 'pressure-test it against real regulatory standards' which has helped the company accelerate its progress 'from idea to a validated MVP (Minimum Viable Product).'
Rounding out the cohort have been two projects; one by Automedica Ltd, investigating the regulatory advantages of using retrieval-augmented generation (RAG) technologies with verified knowledge bases and Large Language Models (LLMs); and the other by health tech startup Newton's Tree testing its Federated AI Monitoring Service (FAMOS) to identify and mitigate AI risks in clinical settings, including performance drift or safety issues.
Results from all four pilot projects will be published later this year, providing valuable insights that will shape the AI Airlock programme moving forward and help inform broader regulatory approaches to the effective and safe use AI in healthcare.
Eligible candidates for the second cohort must demonstrate that their AI-powered medical device has the potential to deliver significant benefits to patients and the NHS, presents a new treatment approach, and offers a regulatory challenge ready to be tested in the Airlock programme.
Applications for cohort two will close on 14 July 2025.
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
BBC News
an hour ago
- BBC News
Botox warning after unlicensed injections seized
A warning has been issued after "cheap" anti-wrinkle and filler injections bought from overseas were seized. Rochdale Borough Council said it had found the unlicensed products - which had no English labelling or instructions - at five business premises in the town where Botox and fillers were being advertised. It said some of the products could not be identified due to poor labelling, posing a potentially "very serious risk of injury". The businesses were referred to the Medicines and Healthcare Products Regulatory Agency (MHRA) which has the power to prosecute. The council said the products were found as part of routine checks to ensure Botox was not being administered to under 18s. It said it could not reveal the names of the businesses at present, and added that there was no evidence found of underage customers being given the cosmetic injections. 'Stay vigilant' Councillor Tricia Ayrton, deputy leader of the council, said: "While there's no issue with wanting to look younger, it's really important you do your homework beforehand with these types of treatment – especially given what our officers found."Know who your prescriber is and who will be administering your treatment. The unlicensed products we've uncovered potentially posed serious health and safety risks, so it's very important you stay vigilant."The council also urged anyone considering cosmetic procedures to check what training and experience providers have and what their insurance toxin, commonly shortened to Botox, is a prescription only product that is regulated by the states customers must first take part in a face-to-face pre-consultation with the prescriber of the Botox product to check their medical history. The prescriber must either be a doctor, dentist, nurse prescriber or pharmacist person administering the injections does not need to be the prescriber, but the prescriber is required to ensure they are properly trained. Listen to the best of BBC Radio Manchester on BBC Sounds and follow BBC Manchester on Facebook, X, and Instagram, and watch BBC North West Tonight on BBC iPlayer.
The Guardian
an hour ago
- The Guardian
Tory peer Liam Booth-Smith takes job at AI firm he encountered at No 10
A former chief of staff to Rishi Sunak who was made a Conservative peer has taken a job as 'external affairs' chief at an artificial intelligence company that he encountered while working at No 10. Liam Booth-Smith, who entered the House of Lords last year, has joined Anthropic, which recently signed a memorandum of understanding about working with the government on AI. His former boss Sunak was a supporter of the UK leading on AI during his time as prime minister, and hosted a meeting where Anthropic was present in May 2023 – which was attended by Booth-Smith as an observer. The Advisory Committee on Business Appointments (Acoba) said Booth-Smith should not engage in any lobbying on behalf of Anthropic for at least two years after leaving the government. Anthropic confirmed to Acoba his role would not involve lobbying and that it would abide by its terms. The committee said that Booth-Smith had not had any role in commercial or contractual decisions relating to Anthropic but noted that No 10 had overall oversight of AI policy. But it added: 'The committee's view is that there remains a risk associated with Lord Booth Smith's influence in government, particularly as he seeks to take up a role in external affairs. Anthropic is currently developing a UK presence to promote its services with a wide range of applications, including with and across the UK government. 'Lord Booth-Smith must avoid engaging in activity that could reasonably be seen as seeking to influence government decisions, including approaching the government, whether special advisers, other officials or ministers about its approach or policy on Anthropic's behalf. If he was to initiate contact with the UK government it would be difficult to manage the risk that this is seen as lobbying.' In addition to a two-year lobbying ban, it advised that Booth-Smith should not draw on any privileged information available to him from his time in government, make use of any contacts made in government to influence policy, or provide advice to Anthropic on any policy he had specific involvement in or responsibility for as chief of staff at No 10 for two years. As a peer, Lord Booth-Smith is already banned from lobbying the government on any subject. Gabe Winn, the chief executive of Blakeney, a lobbying firm, who is campaigning against peers being paid to offer political advice to companies, questioned whether Booth-Smith should be doing the external affairs role for an AI company as well as being a legislator. An Anthropic spokesperson said: 'Liam Booth-Smith is following all applicable Acoba guidance and the House of Lords Code of Conduct in his role.' Booth-Smith has been approached for comment.
The Guardian
4 hours ago
- The Guardian
Doge employee ‘Big Balls' has resigned, says White House official
One of the US so-called 'department of government efficiency' (Doge) service's best-known employees, 19-year-old Edward Coristine, has resigned from the US government, a White House official said on Tuesday, a month after the acrimonious departure of his former boss Elon Musk. The White House official gave no further details on the move and Coristine did not immediately return an email seeking comment. Coristine worked at Musk's brain connectivity company Neuralink before joining the tech billionaire as he led Doge established by the Trump administration earlier this year. Doge has overseen job cuts at almost every federal agency but is starting to see losses itself. Key Musk lieutenant Steve Davis, who was in charge of day-to-day running of Doge, has also left, along with others. The White House has said that Doge's mission will continue. Coristine's youth and online moniker 'Big Balls' became a pop-culture meme as Doge swept through the US government, seizing data and firing employees en masse. Last month, Reuters reported that Coristine was one of two Doge associates promoting the use of AI across the federal bureaucracy. Media outlets, including Wired which first reported his departure, revealed that Coristine had been active in a chat room popular with hackers and previously had been fired from a job following an alleged data leak. In March, Reuters reported that Coristine had provided tech support to a cybercrime gang that had bragged about trafficking in stolen data and harassing an FBI agent. Beginning around 2022, while still in high school, Coristine ran a company called DiamondCDN that provided network services, according to corporate and digital records reviewed by Reuters and interviews with half a dozen former associates. Among its users was a website run by a ring of cybercriminals operating under the name 'EGodly', according to digital records preserved by the internet intelligence firm DomainTools and the online cybersecurity tool The digital records reviewed by Reuters showed the EGodly website, was tied to internet protocol addresses registered to DiamondCDN and other Coristine-owned entities between October 2022 and June 2023, and that some users attempting to access the site around that time would hit a DiamondCDN 'security check'. In 2023, EGodly boasted on its Telegram channel of hijacking phone numbers, breaking into unspecified law enforcement email accounts in Latin America and Eastern Europe, and cryptocurrency theft. Early that year, the group distributed the personal details of an FBI agent who they said was investigating them, circulating his phone number, photographs of his house, and other private details on Telegram. EGodly also posted an audio recording of an obscene prank call made to the agent's phone and a video, shot from the inside of a car, of an unknown party driving by the agent's house in Wilmington, Delaware, at night and screaming out the window: 'EGodly says you're a bitch!' Reuters could not independently verify EGodly's boasts of cybercriminal activity, including its claims to have hijacked phone numbers or infiltrated law enforcement emails. But it was able to authenticate the video by visiting the same Wilmington address and comparing the building to the one in the footage. The FBI agent targeted by EGodly, who is now retired, told Reuters that the group had drawn law enforcement attention because of its connection to swatting, the dangerous practice of making hoax emergency calls to send armed officers swarming targeted addresses. The agent didn't go into detail. Reuters is not identifying him out of concern for further harassment. 'These are bad folks,' the former agent said. 'They're not a pleasant group.'
