AI Airlock is a programme or ‘sandbox’ that is designed to help track the safety and performance of AI tools used in healthcare. It is led by the Medicines & Healthcare products Regulatory Agency (MHRA), the UK’s regulatory body.
There is a global drive to grow the current healthcare landscape by providing more safe and effective methods for testing new drugs and medical products. AI and technology are very much at the forefront of this and promise more efficient and innovative options.
As a part of AI Airlock, there are several ongoing projects, such as:
- Oncoflow, which is a platform that ‘focuses on the explainability of AI in oncology,’ to help provide more specific and in-depth explanations for diagnosis and treatment options
- SmartGuideline, which ‘aims to improve accuracy and trust in AI-driven clinical decision support systems’
- PACS Radiology Auto Impression, which ‘is pioneering efforts to validate AI-generated radiology reports using synthetic data.’ This is to ensure that it is representative of real-world data
These are just a few examples of the various projects that are taking place within the scope of AI Airlock, all with the goal of addressing and improving the validity, explainability and safety of AI in clinical settings.
AI Airlock is also supported by the UK government’s Regulatory Innovation Office (RIO) and is a part of the AI Opportunities Action Plan which aims to increase the number of AI tools and technologies used in healthcare.
It entered the second phase on 23 June 2025 and the MHRA will be providing further information soon.
For more information on the regulatory sandboxes and the AI Airlock project, see also our Regulatory Rapporteur article ‘Sandbox of trust: Regulated AI for health promotion and disease prevention’.
This year’s Symposium takes place on 29 September - 2 October and showcases fascinating sessions on the latest AI technological advancements. Book your place now.