Regulatory Perspectives for AI/ML Implementation in Pharmaceutical GMP Environments

Abstract

Integrating artificial intelligence (AI) and machine learning (ML) into pharmaceutical manufacturing processes holds great promise for enhancing efficiency, product quality, and regulatory compliance. However, implementing good manufacturing practices (GMP) in regulated environments introduces complex challenges related to validation, data integrity, risk management, and regulatory oversight. This review article comprehensively analyzes current regulatory frameworks and guidance for AI/ML in pharmaceutical Good Manufacturing Practice (GMP) settings, identifies gaps and uncertainties, and proposes considerations for future policy development. Emphasis is placed on understanding regulatory expectations across various agencies, including the US FDA, EMA, and MHRA. This article examines verified case studies and pilot programs that demonstrate the successful application of AI/ML under regulatory scrutiny, as well as recent developments in regulatory frameworks and implementation strategies. Ultimately, this article emphasizes the importance of a risk-based life cycle approach and the need for advancements in regulatory science to accommodate the dynamic nature of AI/ML technologies.