Digital and AI-enabled models in pharmaceutical development and manufacturing: a regulatory-focused industry survey

Digital and artificial intelligence (AI)-enabled tools are increasingly being applied across pharmaceutical Chemistry, Manufacturing and Controls (CMC) activities, yet adoption, benefits and regulatory barriers remain incompletely characterised. The Digital CMC Centre of Excellence in Regulatory Science and Innovation (CERSI), led by the University of Strathclyde and supported by the UK Medicines and Healthcare products Regulatory Agency (MHRA), conducted a survey to assess the landscape and inform priorities. A survey was distributed to industry stakeholders, including innovator and generic manufacturers, non-prescription medicine manufacturers, and contract development and manufacturing organisations. Responses were analysed descriptively to evaluate adoption patterns, drivers, barriers, perceptions and support needs. Most respondents reported experience with digital CMC tools, and many reported experiences with AI-enabled tools, across small and large molecule drug substance (DS) and drug product (DP) development. Adoption was more prevalent in small molecule applications, with greater implementation in drug product than drug substance. Fewer than 15% of tools had been included in submissions. Key drivers included process robustness, efficiency, time to market and cost reduction. Principal challenges included culture change, skills gaps and uncertainty regarding regulatory expectations, particularly for AI-enabled applications and evidence requirements for submissions and inspections. Respondents generally agreed that identifiable patient data are not used in Quality/GMDP AI applications, but there was no consensus that current regulatory frameworks fully address AI governance principles. Strong support was expressed for harmonised guidance, training and dialogue. Digital CMC tool adoption is increasing but remains uneven; greater regulatory clarity, harmonisation and collaboration are critical for scalable implementation.