Researchers from the European Medicines Agency (EMA) have outlined key regulatory research priorities for artificial intelligence (AI) throughout the lifecycle of medicines, emphasizing the importance of accuracy and reliability in AI tools.
The EMA's recent preprint paper is based on a survey involving 273 stakeholders, including regulators, industry professionals, and patients. The findings highlight that ensuring the trustworthiness of AI-generated outputs is paramount, particularly as these outputs influence regulatory decisions. Following accuracy and reliability, the concerns of data governance, ethics, and bias prevention were also identified as significant priorities.
The researchers proposed ten priority areas for further investigation, with the top three focusing on maintaining AI system robustness amidst evolving data and the necessity for explainability in AI models. This initiative coincides with the EMA's release of its 2025 AI Observatory report, which discusses AI applications in regulatory submissions and outlines ongoing collaborations aimed at safe AI integration in healthcare.
As the EMA prepares for an upcoming regulatory science research meeting, the emphasis on developing robust guidelines for AI tools reflects a growing recognition of the need for trustworthy and ethical AI applications in the life sciences sector, underscoring the potential for transformative advancements in drug development and patient care.