An experimental bird flu vaccine designed with artificial intelligence fully protected mice and ferrets against multiple virus strains in laboratory tests, raising hopes for a broader approach to preventing future influenza pandemics, according to research published in 2026.
Researchers at the University of Cambridge and its spinout company DIOSynVax developed the vaccine to overcome a major weakness in existing bird flu vaccines, which offer limited protection against different viral variants.
The study, published in Cell Reports Medicine, found that only about half of the animals vaccinated with a conventional vaccine survived exposure to a mixture of virus variants, while all animals given the AI-designed vaccine survived.
The research was led by Professor Jonathan Heeney. Independent laboratories in Canada and Italy also validated the vaccine's effectiveness against viruses collected from different sources.
The research team analyzed genetic information from bird flu viruses identified through global surveillance programs.
Using artificial intelligence, scientists designed special antigens, known as "super-antigens," that combine common features found across multiple bird flu strains.
Unlike conventional vaccines that target a specific viral strain, the new approach aims to provide broader protection against genetically different variants.
Researchers first tested the most promising antigen in laboratory cells before evaluating it in animals. According to Heeney, these specially designed antigens do not naturally exist in viruses.
Co-researcher Sneha Vishwanath said the super-antigens could interfere with several mechanisms involved in viral replication, potentially making it harder for different virus strains to spread.
Many governments stockpile bird flu vaccines for potential human use. These vaccines are developed using selected virus samples through the World Health Organization's pandemic influenza preparedness program.
However, the Cambridge study found that conventional vaccines targeting a specific strain may provide limited protection against other variants.
This creates a challenge for governments because influenza viruses continue to mutate. As new variants emerge, existing vaccines may need frequent updates, increasing the financial and logistical burden of pandemic preparedness.
The AI-designed approach could offer a different strategy by targeting features shared across multiple strains rather than focusing on a single variant.
Despite the promising animal results, researchers caution that the vaccine has not yet been tested in human clinical trials.
Scientists cannot use successful experiments in mice and ferrets as proof that the vaccine is safe or effective in people.
A coronavirus vaccine developed using the same AI-based methodology reportedly demonstrated safety in human trials in 2026.
However, the bird flu vaccine must undergo its own clinical testing before researchers can establish its safety and effectiveness in humans.
Bird flu viruses were reported in Australia and New Zealand for the first time in 2026, adding to concerns about the virus's global spread.
Researchers are also monitoring pigs because their respiratory systems contain receptors similar to those found in humans. This similarity raises concerns about the possibility of viral transmission between species.
Heeney highlighted the importance of monitoring this risk as scientists work to develop broader protection against emerging influenza strains.
Although the AI-designed vaccine has shown encouraging results in animal experiments, further research and human trials will be essential to determine whether it can become a reliable tool against future bird flu outbreaks.
