In an interview with News Medical-Life Sciences, James McDonagh, Principal AI Scientist at ApconiX, discussed how modern AI approaches are reshaping nonclinical safety assessment. Drawing on experience spanning computational chemistry, machine learning, pharmaceutical research, and safety science, McDonagh noted that AI and machine learning have been used in early discovery for many years, but the biggest recent shift is in areas dominated by human processes and communication, as computers can now interact directly with human language and analytical images. He stated that in nonclinical safety, the biggest impact will be AI working alongside safety scientists to unearth key data, connect safety science with earlier discovery and later clinical work, and help process, predict, and summarize important outcomes. Regarding data-driven property prediction, McDonagh explained that data-driven models like Quantitative Structure Activity Relationships offer rapid methods for predicting molecular properties and can be very accurate for new molecules in related regions of chemical space to the training data, but they typically do not generalize across large diverse chemical and biological spaces, requiring real care when applied to novel chemistry.