Stakeholders need AI tools tailored for animal to human drug research

Making Sense of Animal-to-Human Drug Development Evidence: Stakeholder Practices, Challenges, and Requirements for AI Tools

Human-Computer Interaction

Summary

Studying how drugs work in animals to predict effects in humans is hard because the evidence is incomplete and people try to use it in different ways. The authors talked with 13 experts involved in this process to learn what challenges they face and how AI could help. They found that while AI can help find and organize studies, people want to be cautious about AI making judgments or interpretations without human checks. The experts want AI tools that are clear about where information comes from, explain uncertainty, and let people stay in control. The authors suggest designing AI that fits different people's roles to make drug research smoother and more transparent.

What this means in practice

  • For pharmaceutical research teams: Use AI tools designed for specific roles to assist in locating, screening, and extracting animal-to-human drug evidence for better decision making.
  • For regulatory review teams: Employ AI systems that maintain transparent source tracking and uncertainty details to support regulatory evaluation of animal-to-human evidence in drug approvals.
  • For medical data platform developers: Develop role-sensitive AI software that integrates transparency and human oversight features to improve synthesis of preclinical and clinical evidence from drug studies.$Commercial implications: Role-sensitive AI tools for drug evidence synthesis can be sold to pharmaceutical companies and regulators to improve efficiency and trust.

Authors

Rosni Vasu, Simona E. Doneva, Benjamin V. Ineichen

Abstract

Animal models are widely used to study human biology and health interventions, yet translating findings from animal studies to humans remains challenging. Evidence across preclinical and clinical research informs experimental and translational decisions. Artificial Intelligence (AI) tools are increasingly reshaping how this evidence is searched, synthesized, and used, but it remains unclear how they should support the diverse stakeholders involved in assessing animal-to-human evidence. We conducted semi-structured interviews with 13 stakeholders to examine their evidence practices, challenges, and expectations for AI support. We found that stakeholders approach the same incomplete evidence base with different goals, expertise, and heuristics. Participants valued AI particularly for locating, screening, and extracting evidence, but were more cautious about automated interpretation and quality judgments. They emphasized transparency, source traceability, uncertainty communication, and human oversight. Based on these findings, we derive design implications for role-sensitive AI tools that support more systematic and transparent reasoning about animal-to-human translation.
View PDFOpen arXiv