How ready are we to use artificial intelligence in our fight against antimicrobial resistance? An ESGAID and EAAS perspective

Comprehensive Summary

This study, presented by Giacobbe and colleagues, examines the expanding role of artificial intelligence (AI) in addressing antimicrobial resistance (AMR), a major global health threat. The authors present a perspective-based analysis synthesizing current literature on AI applications across the antimicrobial lifecycle, including antibiotic discovery, laboratory surveillance, diagnostic support, and antimicrobial stewardship. Rather than conducting an empirical study, the paper reviews emerging technologies and evaluates both their capabilities and implementation challenges within clinical and public health systems. The authors find that AI has demonstrated promise in accelerating antibiotic discovery through predictive modeling, improving surveillance by identifying resistance patterns in large datasets, enhancing diagnostic accuracy for resistant infections, and supporting clinician prescribing decisions. However, substantial barriers limit widespread adoption. Key concerns include trade-offs between model accuracy and explainability, inconsistent data quality, limited digital infrastructure in certain healthcare systems, and insufficient transparency in training datasets. Ethical and regulatory considerations—such as accountability, bias, and liability—further complicate integration. In the discussion, the authors emphasize that AI should augment rather than replace clinician judgment, and that multidisciplinary collaboration is essential to ensure responsible and equitable deployment in antimicrobial stewardship efforts.

Outcomes and Implications

Antimicrobial resistance threatens the effectiveness of modern medicine, increasing morbidity, mortality, and healthcare costs worldwide. Traditional drug development pipelines and stewardship strategies alone may be insufficient to keep pace with emerging resistant pathogens, necessitating innovative and scalable solutions. Clinically, AI tools could significantly improve early detection of resistant infections, guide targeted antibiotic prescribing, and reduce inappropriate antimicrobial use—key drivers of resistance. Predictive models may help clinicians select optimal empiric therapy while awaiting culture results, potentially improving patient outcomes and reducing unnecessary broad-spectrum antibiotic exposure. However, safe implementation requires high-quality datasets, clear regulatory frameworks, and clinician training to interpret AI-generated recommendations appropriately. The authors suggest that near-term implementation is most feasible in surveillance and decision-support applications within well-resourced healthcare systems, while broader integration into antibiotic discovery and global stewardship efforts will require sustained regulatory and infrastructural development. Ultimately, AI-enhanced stewardship programs could strengthen global responses to AMR, provided they are implemented with transparency, accountability, and strong human oversight.