Beyond ADMET: AI for medicines safety - survey

AI has the potential to improve medicines safety evaluation, including the prediction and assessment of a medicine’s absorption, distribution, metabolism, excretion and toxicity (ADMET), by integrating traditional preclinical data with clinical and real-world data which reflects the variability across populations, disease states and clinical conditions. This could provide new ways to predict how medicines behave in the body and perform in real-world settings, across different populations and clinical conditions, and help identify potential safety risks earlier.

We are interested in understanding how organisations are currently using, developing or planning to use AI based approaches, as well as the barriers or challenges they face. This includes areas such as model development, validation, access to data, data sharing and regulatory considerations.

The findings will inform the development of the ‘Beyond ADMET: AI for medicines safety’ regulatory sandbox, announced in June 2026, and support future regulatory work in this area. They will also help shape future engagement with innovators and researchers, including workshops and other opportunities to explore key issues in more detail.

We welcome views from organisations involved across the medicines development pathway, including toxicology, preclinical and clinical research and testing, as well as life sciences, biotechnology, AI and other related sectors.

Please share perspectives and experiences that are relevant to your organisation. Where possible, provide examples to help us understand current practice, challenges and priorities.

You do not need to be using AI for medicines safety-related approaches today to respond. We are also interested in future plans and areas of unmet need.