Trustworthy and explainable AI for drug discovery

Taylor & Francis Online
July 6, 2026

AI/ML is listed as a top investment priority by ~60% of R&D leaders and major biopharmas everywhere are pursuing major AI initiatives. However, an intricate, high stakes, and regulated field such as drug discovery and development requires a careful approach towards AI. 

This Comment discusses the power and potential of explainable AI (XAI) to both (1) engender greater trust in results, by providing a “biological sanity check” of AI tools, and (2) generate novel insights in areas ranging from chemical exploration to biomarker discovery. Beyond this, we emphasize the necessity of high data quality and proper annotation for reliable AI-based predictions, as well as ways to achieve this through AI-based automation of analysis. 

Finally, we underline the importance of choosing the appropriate applications and AI approach (machine-learning based classification versus LLM, using human-in-the-loop, etc.) when embarking on the integration of AI technology.