Artificial Intelligence (AI) is rapidly becoming an important tool across the pharmaceutical and biotechnology sectors, with applications ranging from process monitoring and predictive analytics to deviation detection and classification. As AI adoption accelerates by life sciences companies, regulators are increasingly focused on ensuring AI technologies are implemented in a controlled, effective, and GMP-compliant manner.
The European Commission has recently issued a draft version of EU GMP Annex 22, a new guideline specifically addressing the use of AI in GMP-regulated manufacturing environments. The proposed annex is intended to complement existing requirements under Annex 11 for computerised systems and introduces a risk-based framework for AI deployment in pharmaceutical manufacturing.
The draft guidance applies to AI and machine learning (AI/ML) systems used in critical GMP applications that may impact patient safety, product quality, or data integrity. The current draft guideline focuses on ‘static’ AI models; models that do not adapt their performance during use by incorporating new data. The draft Annex 22 states that ‘dynamic’ models that continuously learn and adapt during use are outside its scope and should not be used in critical GMP applications.
Several key themes emerge throughout Annex 22:
- Clearly defining the intended use of AI models
- Robust validation and acceptance criteria
- Independence and quality of test datasets
- Explainability and transparency of AI outputs
- Human oversight and governance
- Ongoing monitoring and lifecycle management
The guidance also highlights the importance of cross-functional collaboration between quality, manufacturing, IT, process SMEs, and data scientists throughout the AI lifecycle.
For companies operating in the cell and gene therapy sector, these developments are particularly significant. Advanced therapy manufacturing generates large and complex datasets, making AI an increasingly attractive tool for process optimisation, analytics, and operational efficiency. However, the regulatory expectations surrounding validation, data governance, and explainability will require careful consideration as AI adoption expands.
While Annex 22 remains in draft form, it provides a strong indication of the future regulatory direction for AI within GMP manufacturing environments. Organisations that begin establishing appropriate governance frameworks, validation strategies, and quality oversight processes today will be better positioned for future compliance and successful implementation of AI-enabled technologies.
HiTech Health is a leading provider of QP services in the EU and the UK. Our QP team have extensive knowledge of pharmaceuticals, biologics and Advanced Therapeutic Medicinal Products (ATMPs) including cell and gene therapies. We can help with all quality and compliance activities from your overall quality management system through quality audits and QMS remediation support.
Contact our team today and schedule a call to learn more about our services for the EU and UK markets.
References: