Friday, November 6, 2026

News

Cambridge AI-Designed Bird Flu Vaccine Outperforms Current Stockpile in Animal Tests

ResearchPatryk Raba

University of Cambridge researchers used algorithms to design a vaccine covering multiple bird flu variants. In mice and ferrets it gave full protection, while the current WHO vaccine yielded about 50% survival.

Contents
  1. What the study showed
  2. How the antigen design works
  3. Why it matters
  4. What comes next

A team at the University of Cambridge has described a bird flu vaccine whose antigens were designed with the help of AI algorithms. In tests on mice and ferrets, it protected the animals against several genetically distinct variants of the virus, while the current WHO vaccine gave about 50% survival. The results so far apply only to animals.

What the study showed

According to Euronews, the work was led by Jonathan Heeney of the Lab of Viral Zoonotics at Cambridge, in collaboration with DIOSynVax Ltd, a University of Cambridge company. The vaccine is meant to target many bird flu strains in a single dose, rather than training the immune system to recognize a single variant.

In the experiments, mice and ferrets were exposed to several genetically different variants of the virus. According to the source, all of these variants have already infected humans and carry potential pandemic risk. Animals given the new vaccine were fully protected, while those immunized with the current WHO preparation survived in about half of cases.

The vaccine also gave broader protection and better survival than the standard single-strain candidates that are kept in stockpiles in case of a pandemic.

How the antigen design works

The researchers collected genetic sequences of bird flu viruses recorded by surveillance programs worldwide. The samples came from wildlife, livestock and infected humans.

Using this data, computational and evolutionary algorithms designed five AI-optimized synthetic antigens. They are meant to reflect the diversity of viruses worldwide and make it harder for the virus to escape immune protection. Heeney stresses that AI greatly speeds up design, but the candidates then have to be checked in the lab for robustness and safety.

It’s designed to anticipate viral evolution, rather than reacting to it - Jonathan Heeney, University of Cambridge

Why it matters

Bird flu is an RNA virus, like the coronaviruses responsible for COVID-19. Such viruses mutate quickly, can jump from animals to humans and can evade immunity and existing vaccines. Current preparedness relies on stockpiling vaccines against variants that have already emerged.

The authors suggest the new approach could protect even against variants that have not yet appeared. That is the researchers' expectation, however, not a result shown in these experiments. The authors call next-generation vaccines a breakthrough for national biosecurity, which is their own assessment.

What comes next

The team plans human clinical trials so that the vaccine could be available in the event of an outbreak. It also plans to extend the method to seasonal flu vaccines and other rapidly evolving RNA viruses.

Earlier in 2026, according to Euronews, the same team produced the first fully AI-designed vaccine to be tested in humans, aimed at viruses such as Ebola and coronaviruses. Until clinical trials are done, there is no data on the efficacy and safety of the new bird flu vaccine in humans.

Share: