A Decade Of AI-Accelerated Drug Discovery Against Antimicrobial Resistance (2015–2025): Insights And Future Directions

Abstract

Antimicrobial resistance (AMR) presents a critical global health threat, demanding innovative approaches beyond conventional antibiotic development. Over the past decade, Artificial Intelligence (AI) has emerged as a transformative tool in addressing AMR by facilitating antibiotic discovery, predictive resistance modelling, diagnostics, and data management. This systematic review synthesized literature from 2015 to 2025 across five scholarly databases PubMed, Scopus, Web of Science, IEEE Xplore, and ScienceDirect using predefined Boolean search strings combining terms such as Artificial Intelligence machine Learning Antimicrobial Resistance, "Drug Discovery", "Diagnostics", and "Surveillance". Out of 372 initially identified articles, 52 met inclusion criteria for relevance, novelty, and methodological robustness.

Key insights reveal that AI-enhanced drug discovery has accelerated the identification of novel antimicrobial compounds and enabled drug repurposing with greater precision. Machine learning algorithms have improved predictive models for resistance patterns, facilitating early intervention and surveillance. AI-driven diagnostic platforms, particularly deep learning-based imaging and decision support systems, demonstrated improved diagnostic accuracy and faster turnaround times, especially in resource-limited settings. However, data challenges, algorithmic biases, and lack of integration with real-world healthcare infrastructure remain critical barriers.

Thematic analysis revealed five dominant themes namely (1) drug discovery and repurposing, (2) diagnostics and decision support, (3) resistance prediction and surveillance, (4) data management and integration, and (5) ethical and regulatory constraints. While thematic convergence supports AI's pivotal role in AMR mitigation, contradictions were evident in reproducibility, interpretability, and translational applicability across diverse health systems.

Future directions call for the development of transparent AI frameworks, stronger cross-disciplinary collaborations, standardized datasets, and policy support to enable AI translation into clinical and public health interventions. Furthermore, integrating AI with genomics, One Health approaches, and mobile-based surveillance systems may significantly enhance AMR response in both high- and low-resource settings.