Artificial intelligence (AI) continues to redefine the future of healthcare, from detecting cancer faster to improving diagnostic accuracy across hospitals. The UK’s Medicines and Healthcare products Regulatory Agency (MHRA) has announced the second phase of its AI Airlock program, a new initiative that provides a controlled environment to test and validate AI-powered medical technologies safely before they reach patients.
Seven AI innovators have been selected to participate in this phase, developing technologies that span advanced cancer diagnostics, eye disease detection, automated clinical note taking, and blood test interpretation. The goal is to help clinicians make faster, more informed decisions while ensuring the highest standards of patient safety and regulatory compliance.
The Airlock acts as a “sandbox”, a space where developers and regulators collaborate to test real-world AI applications, measure potential risks, and explore clear pathways to approval. Insights from this phase will also guide the MHRA’s National Commission into the Regulation of AI in Healthcare, which brings together experts, clinicians, and patient representatives to shape the future of AI oversight in medicine.
Health Innovation Minister Zubir Ahmed described the initiative as a milestone in the government’s mission to make the NHS the most AI-enabled healthcare system globally. “The AI revolution is here,” he said. “The Airlock ensures innovations are tested rigorously while moving at pace toward transforming care delivery.”
Science Minister Lord Vallance added that AI in healthcare not only improves patient outcomes but also strengthens the UK’s economy by supporting the life sciences sector and reducing productivity losses linked to poor health.
According to Lawrence Tallon, MHRA Chief Executive, this program cements the UK’s leadership in responsible AI regulation. “We’re the first country to create a dedicated regulatory sandbox for AI medical devices,” he said. “The collaboration between innovators and regulators is shaping the next generation of safe, effective AI tools for healthcare.”
Phase two builds on the success of the pilot phase, which worked with companies including Philips, Newton’s Tree, OncoFlow, and Automedica Ltd. These early partnerships helped identify key improvements in areas like synthetic data validation, AI decision explainability, and mitigation of AI hallucinations challenges central to ensuring safe adoption of emerging technologies.
Learn more: AI tools that could detect diseases earlier selected for next phase of MHRA’s ‘AI Airlock’ programme
