Regina Barzilay and Jim Collins Win Pritzker Prize for AI Antibiotic Discovery
MIT researchers honored for pioneering AI-driven methods to identify novel, life-saving antibacterial candidates.
1Pritzker Prize Recognition
Regina Barzilay and James Collins received the 2026 Pritzker Prize for AI in Science Research Excellence. The award honors their collaborative work at MIT, which integrates deep learning, bioengineering, and experimental validation. Their research focuses on identifying novel antibacterial candidates to combat the global crisis of drug-resistant pathogens and superbugs.
2Breakthrough With Halicin
In 2020, the team published a landmark paper in Cell describing halicin, the first antibiotic discovered by AI. Trained on 2,500 molecules, the deep learning model identified this broad-spectrum candidate in weeks rather than years. Halicin effectively targets resistant bacteria like Clostridium difficile, Mycobacterium tuberculosis, and Acinetobacter baumannii during laboratory testing.
3Targeted Abaucin Discovery
Following their initial success, the researchers identified abaucin in 2023. Unlike broad-spectrum predecessors, abaucin acts as a narrow-spectrum antibiotic specifically targeting A. baumannii. This precision minimizes the risk of bacteria developing resistance while protecting beneficial gut flora, marking a significant advancement in how scientists approach the design of targeted medical treatments.
4Shifting Discovery Paradigms
James Collins notes that antibiotic discovery has historically relied on slow, costly, and unpredictable trial-and-error strategies. The integration of generative AI represents a fundamental shift, moving from simple discovery tools toward sophisticated design platforms. This evolution allows researchers to navigate complex molecular spaces that were previously inaccessible through traditional laboratory screening methods.
5The Biological Modeling Challenge
Regina Barzilay emphasizes that despite recent progress, drug discovery remains in its early stages. Current AI capabilities automate specific segments of the pipeline rather than the entire process. To achieve a total transformation, the field must prioritize building more realistic and comprehensive AI models that accurately represent the complexities of human biology.
6Institutional Collaborative Framework
The award-winning research was conducted through the Abdul Latif Jameel Clinic for Machine Learning in Health. Barzilay and Collins serve as faculty co-leads, working alongside affiliates from the MIT Schwarzman College of Computing and the School of Engineering. This interdisciplinary environment facilitates the convergence of computational power and biological engineering expertise.
7Historical Research Context
The collaboration began with Jonathan Stokes, a former MIT postdoc who served as the first author on their 2020 breakthrough paper. Stokes, now an assistant professor at McMaster University, continues to work with Barzilay and Collins. Their ongoing partnership highlights the importance of sustained academic collaboration in translating theoretical AI models into tangible medicine.
8Future of Medical Innovation
The Pritzker Prize underscores the growing necessity of AI in addressing antimicrobial resistance. By reducing the time required for candidate identification, these methods provide a scalable solution to a global health threat. Future developments will likely depend on refining biological simulations to ensure that AI-designed drugs are both safe and effective for patients.