An international team of researchers from Inria Saclay (France) and the Indraprastha Institute of Information Technology Delhi (IIIT-Delhi, India) has developed a groundbreaking artificial intelligence (AI) technique aimed at combating drug-resistant infections. The system proposes alternative antibiotic treatments using existing drugs—offering a powerful tool in the global battle against antimicrobial resistance (AMR).
Antimicrobial resistance is escalating into one of the most pressing public health concerns worldwide. It emerges when bacteria adapt in ways that render common antibiotics ineffective. This growing threat means that everyday infections such as pneumonia, urinary tract infections, or even minor cuts could become deadly. The issue is especially critical in low- and middle-income nations, where over 70% of hospital-acquired infections show resistance to at least one frequently used antibiotic.
Developing new antibiotics remains a slow, expensive process—often taking more than ten years and hundreds of millions of dollars to bring a single new drug to market. As a result, many researchers and clinicians are turning to drug repositioning, a more economical strategy that seeks new applications for medications already in circulation.
In response to this challenge, a research team led by Dr. Emilie Chouzenoux of Inria Saclay and Dr. Angshul Majumdar of IIIT-Delhi has introduced a sophisticated machine learning model capable of recommending effective alternative antibiotics for resistant infections. The project team also includes research engineer Stuti Jain and graduate students Kriti Kumar and Sayantika Chatterjee.
This AI-driven method stands out for its hybrid design. Rather than depending solely on predefined rules or narrow datasets, the algorithm learns from actual clinical treatment data. The researchers assembled comprehensive antibiotic usage guidelines from top Indian hospitals, capturing how physicians manage infections in real-world scenarios. This information was merged with molecular data—including bacterial genomes and the chemical structures of antibiotics—to identify overlooked treatment possibilities.
The AI system was evaluated through case studies involving several dangerous multi-drug resistant bacteria:
- Klebsiella Pneumoniae – a primary cause of ventilator-associated pneumonia and bloodstream infections;
- Neisseria Gonorrhoeae – responsible for gonorrhea, which has developed resistance to standard treatments;
- Mycobacterium Tuberculosis – the bacterium behind tuberculosis, a major health issue in countries such as India.
In these tests, the model successfully identified antibiotics that were either known to work or showed promising potential for repurposing. These findings were validated through resistance data and expert assessments, reinforcing the system’s clinical relevance.
The implications of this research extend beyond individual cases. By offering a scalable, data-informed support system for medical professionals and policymakers, the AI tool could be integrated into hospital workflows or national health strategies. Its use could help reduce delays in treatment, support responsible antibiotic use, and ultimately save lives.
“This collaboration showcases the power of AI and international research in addressing urgent healthcare challenges,” said Dr. Majumdar. “Our approach leverages existing knowledge to make better treatment decisions and provides a quicker path to addressing AMR.”
The team envisions this tool being adopted widely, particularly in regions with limited diagnostic infrastructure, as part of standard infection management protocols.