Scientists at Stanford School of Engineering have developed a new class of bacteria-killing materials known as polymers, which could serve as a promising alternative to traditional antibiotics. These polymers are designed to mimic antimicrobial peptides—small proteins naturally present in human bodily fluids that can destroy bacteria on contact without allowing the bacteria to become resistant. Unlike conventional antibiotics, which often need to enter the bacterial cell or target specific proteins, these peptides work by disrupting the bacteria's cell membrane, a method that makes resistance development less likely. To identify polymers with antimicrobial properties similar to those of peptides, the researchers employed an artificial intelligence (AI) model. This allowed them to analyze millions of potential polymer candidates and find those with similar characteristics. However, the team encountered a challenge due to the limited availability of data on antimicrobial polymers. To overcome this, they trained the AI using the chemical properties of known antimicrobial peptides and applied that knowledge to the polymers. To tackle the issue of data scarcity, the researchers used a strategy based on uncertainty. They focused on the polymers that generated the most disagreement among different AI models and synthesized the 20 most disputed candidates. After testing these candidates and feeding the results back into the AI models, the system became better at identifying effective polymers. The researchers eventually narrowed their list to 10 polymer candidates, which were synthesized and tested against E. coli bacteria. All 10 performed significantly better than expected, with one showing strong effectiveness against biofilms—thick layers of bacteria that are especially difficult to treat. The polymers were also effective against Gram-positive bacteria such as Staphylococcus aureus. The team believes this approach could lead to the development of drugs that specifically target harmful bacteria without affecting human cells or beneficial bacteria. The findings of this study were published in the journal Matter.