At the Pharma Meets AI conference in Barcelona, Spain, April 2026, experts discussed the rapid adoption of AI in drug development and the emerging challenges of trust and governance. Despite advancements in predictive modelling and automation, issues concerning quality, bias, and model reliability hinder widespread deployment. Dr. Debarshi Dey from Galapagos noted that AI must transition from experimental use to integration within decision-making frameworks. AI is making significant impacts in prediction, personalisation, and productivity, yet trust is crucial, especially in high-stakes areas like drug discovery. Ensuring AI models are trained on high-quality, representative datasets is essential to avoid biases that could lead to misleading predictions. Consequently, there is a growing emphasis on robust validation, clear usage context, and continuous monitoring of AI performance. Regulatory bodies are shifting towards active enablement of AI, focusing on auditability, transparency, and reproducibility, reflecting a broader industry trend towards viewing AI as a continuously governed system. As AI adoption progresses, building trust through strong governance will be vital for its evolution from experimental tools to a core part of decision-making in drug development.
What are the key barriers to scaling AI in drug development?
