Artificial intelligence–driven pharmacovigilance: transforming signal detection, real-world drug safety monitoring, and regulatory decision-making

INTRODUCTION: Pharmacovigilance is undergoing rapid transformation due to the expansion of digital healthcare systems, biomedical data generation, and pharmaceutical innovation. Conventional drug-safety monitoring methods remain limited by underreporting, delayed signal identification, fragmented datasets, and restricted predictive capability. Artificial intelligence (AI) has emerged as a promising approach for improving signal detection, real-world drug-safety surveillance, and regulatory decision-making. AREAS COVERED: This review evaluates application of AI technologies in modern pharmacovigilance and regulatory science. It discusses use of machine learning, deep learning, neural networks, and natural language processing for adverse-event detection, literature mining, signal prioritization, and analysis of electronic health records (EHR) and social-media data. Emerging technologies, including large language models, federated learning, blockchain, digital twins, and Internet of Things (IoT)-enabled monitoring systems, are also explored. Literature was identified through structured searches of MEDLINE/PubMed, Scopus, Web of Science, Embase, and Google Scholar, covering publications from January 2000 to February 2026. EXPERT OPINION: AI-driven pharmacovigilance has potential to improve real-time safety surveillance, predictive risk assessment, and automated regulatory workflows. However, broader implementation remains challenged by concerns related to interpretability, data quality, cybersecurity, algorithmic bias, and regulatory validation. Future progress will depend on transparent, explainable, and human-supervised AI systems integrated within globally interoperable healthcare infrastructures.