Lawrence Tallon
DesignationChief Executive of the Medicines and Healthcare products Regulatory Agency (MHRA)
At the UKAI and Curia annual Healthcare and Life Sciences Parliamentary Showcase, Chief Executive of the Medicines and Healthcare products Regulatory Agency (MHRA), Lawrence Tallon set out a clear message to Parliament, AI Healthcare Companies, academia and the NHS – regulation should not be seen as a barrier to innovation, but as one of the conditions that enables it.
Speaking as keynote speaker at the event, Tallon argued that the UK has the scientific, technological, and biomedical strengths needed to improve patient care, and drive economic growth. However, he warned that those strengths will only translate into better outcomes if proven innovation can reach clinicians and patients more quickly.
For Tallon, he clearly articulated that the UK can innovate effectively from a world-leading position. What he questioned is whether the system can adopt innovation at scale.
AI Healthcare Companies: Regulation and innovation should not be opposites
Tallon opened by challenging the idea that regulation and innovation naturally sit in tension. He said that while the two words are “not always” heard in the same sentence, he believes regulation can be a catalyst for innovation.
Drawing on his own experience in public service innovation before joining the MHRA, he pointed to his work as Deputy Chief Executive at Guys and St Thomas’ NHS Foundation Trust on the first medical use case of drone technology in the UK. That example, he suggested, shows that public services can use new technology both to improve delivery and support wider economic growth.
If the UK wants to improve healthcare outcomes and grow its life sciences, medtech and artificial intelligence (AI) sectors, it needs regulation that gives innovators a clear, trusted and proportionate route to market.
The current AI pathway is not enough for the next generation of tools
Tallon acknowledged that there is already an established regulatory route for AI in healthcare. However, he said this has largely been designed around the dominant AI use case of recent years – image recognition, particularly in radiology.
In that context, AI systems are often assessed on whether they can read scans as well as, or better than, humans. But Tallon warned that the new era of AI will involve far more complex products, including large language models and large medical language models, with a much wider range of possible clinical and operational applications.
That shift, he argued, demands a regulatory framework that is fit for the next generation of AI, rather than one built only around earlier use cases.
“We must weigh the risk of inaction just as much as the risk of action.”
A national process for building trust
Tallon highlighted the Government’s National Commission on the Regulation of AI in Healthcare, which is bringing together around 100 experts from big tech, small tech, government, frontline clinicians, academics, patients and patient representatives.
He said the Commission has already received 700 responses to its call for evidence and held 20 events across the country to hear directly from patients and citizens about what they want from AI, and what concerns them.
That engagement matters because trust will be central to whether AI can be adopted safely and confidently across the NHS. For Tallon, the answer to complex and adaptive technologies is not to regulate behind closed doors, but to build an expert, authoritative and inclusive process.
The risk of inaction must count too
One of Tallon’s strongest messages was that regulators must consider not only the risk of approving a product, but also the risk of failing to make useful tools available.
He said it is easy for a regulator to ask what might happen if a product is approved and something goes wrong. But he said regulators must also think about clinicians and patients who could benefit from technologies that are already proven to be better than current practice.
Where AI is “good enough” and demonstrably better than what is currently available, Tallon argued that the system should be more ambitious and more permissive in making those tools available to clinicians and patients.
From a high jump to a hurdles race
Tallon’s second key message was that AI requires a fundamentally different regulatory model from traditional medicines regulation.
Medicines regulation has historically involved a very high barrier before a product reaches patients. But AI, he argued, is different. Because it is complex, adaptive and capable of changing over time, the regulatory system should place greater emphasis on post-market monitoring, real-world evidence and iterative assessment.
He compared this approach to a hurdles race rather than a high jump: not one almost impossibly high barrier to entry, but a series of proportionate steps that allow products to reach patients more quickly while continuing to monitor performance in real-world NHS settings.
The five Pros of AI regulation
Tallon’s final key message was that AI products are context-specific. Regulators cannot only look at the device itself; they also need to consider the system in which it is used.
He set out the MHRA’s five Pros model: the product, the producer, the provider, the professional and proportionality.
“Think less about a high jump and more about a hurdles race.”
The product must be safe and perform as expected. The producer must behave ethically, report safety signals and maintain the model after deployment. The provider must understand how to use AI in clinical workflows, manage bias and understand error. The professional must understand the intended effect of the algorithm in the same way a clinician understands the intended effect of a medicine. Finally, the overall system must be proportionate to risk.
A more ambitious route to adoption
Tallon concluded that the UK has an opportunity to design an optimal model for regulating AI in healthcare, with the Commission’s initial report due to be published soon with the final report coming later in the year.
His central message was that smarter regulation can help the UK get proven technology to patients and clinicians more quickly, while maintaining safety, trust and proportionality.
For a health system under pressure, and a life sciences sector with major growth potential, that balance could be crucial.
Get your new edition of ChamberUK. Our parliamentary journal.
Photo Credit: The wub with ChatGPT overlay