Asian Journal of Advances in Medical Science

Antimicrobial resistance (AMR) represents a growing global health threat, driven by the widespread misuse of antibiotics and the limitations of current stewardship strategies. Conventional approaches to antimicrobial stewardship rely heavily on empirical prescribing and delayed microbiological confirmation, often resulting in suboptimal antibiotic selection and contributing to acceleration of resistance. In this context, artificial intelligence (AI) has emerged as a promising tool to enhance decision-making through predictive and data-driven approaches.

This review explores the role of AI in antimicrobial stewardship, with a focus on its application in predicting antimicrobial resistance, guiding antibiotic selection, and supporting clinical decision-making through integrated systems. Machine learning and deep learning models have demonstrated the ability to analyse complex, multidimensional datasets, including electronic health records, microbiological data, genomic information, and antibiotic usage patterns, to generate early and accurate predictions. These capabilities enable a shift from reactive to predictive care, particularly in high-risk settings such as intensive care units and sepsis management.

Despite these advancements, significant challenges remain. Limitations related to data quality, model generalizability, interpretability, and integration into clinical workflows continue to hinder widespread adoption. Additional barriers include clinician trust, infrastructure constraints, and ethical considerations such as data privacy and accountability. These challenges are further amplified in low-resource settings, where digital infrastructure and access to microbiological diagnostics may be limited.

Future progress will depend on the development of robust datasets, standardized reporting frameworks, and seamless integration of AI tools into healthcare systems. With appropriate validation and implementation, AI has the potential to transform antimicrobial stewardship by enabling more precise, timely, and context-aware antibiotic use. Ultimately, the integration of AI into clinical practice represents a critical step toward addressing the global burden of antimicrobial resistance.