Abstract
Antimicrobial resistance (AMR) is a critical global health threat and artificial intelligence (AI) presents new opportunities for our response. However, research priorities at the AI-AMR intersection remain undefined. This study aimed to identify and prioritise key areas for future investigation. Using a modified James Lind Alliance approach, we conducted semi-structured interviews with eight experts in AI and AMR between February and June 2024. Analysis of 338 coded responses revealed 44 distinct themes. Major barriers included fragmented data access, integration challenges and economic disincentives. The top ten priorities identified were: Combination Therapy, Novel Therapeutics, Data Acquisition, AMR Public Health Policy, Prioritisation, Economic Resource Allocation, Diagnostics, Modelling Microbial Evolution, AMR Prediction and Surveillance. A notable limitation was the underrepresentation of data from high-burden regions, limiting the generalisability of findings. To address these gaps, we propose the novel BARDI framework: Brokered Data-sharing, AI-driven Modelling, Rapid Diagnostics, Drug Discovery and Integrated Economic Prevention.