Researchers at the University of Cambridge say they have developed a “fundamentally new” type of vaccine approach using artificial intelligence, marking what they describe as the first instance of an AI-designed vaccine component being tested in human trials.
The experimental vaccine is designed to target entire families of viruses, including coronaviruses responsible for Covid-19 and related strains found in animals that could potentially trigger future outbreaks. Scientists behind the work say the aim is to move beyond vaccines that need frequent updates and instead create broader protection against evolving pathogens.
Traditional vaccines are typically developed using a single circulating strain of a virus. However, many viruses mutate rapidly, reducing vaccine effectiveness over time and requiring regular reformulation, as seen with seasonal influenza and Covid-19 boosters.
In this study, Cambridge researchers used genetic data from multiple known coronaviruses collected through global surveillance programmes. Artificial intelligence was then used to analyse these genetic sequences and design a synthetic “super-antigen” intended to train the immune system to recognise a wide range of related viruses, even if they change or cross from animals to humans.
Antigens are the key elements in vaccines that trigger immune responses. By designing a broader antigen, researchers aim to prepare the immune system for future variants that have not yet emerged.
Professor Jonathan Heeney of the University of Cambridge said the approach represents a shift in how vaccines could be developed in the future. “We’re always behind,” he said, adding that the goal is to anticipate outbreaks rather than react to them. He described the use of AI in this context as “surprising” and said it could reshape pandemic preparedness.
Early-stage human trials involved 39 volunteers to assess safety. A second phase involving around 200 participants is planned to evaluate immune response effectiveness. Findings published in the Journal of Infection show only a modest immune response so far, but researchers say the results remain promising for such an early-stage technology.
Professor Saul Faust of the University of Southampton, who helped conduct the trials, said the AI approach has clear potential, particularly in addressing rapidly evolving viruses. He noted that designing vaccines for future pandemic threats is an area where artificial intelligence may offer significant advantages.
The Cambridge team is also expanding research into universal flu vaccines, H5N1 bird flu protection, and vaccines targeting viral haemorrhagic fevers such as Ebola. Some of these diseases currently have limited or no effective vaccines available.
Experts not involved in the study, including Professor Andy Pollard of the Oxford Vaccine Group, have described the early results as encouraging but stressed that human immune responses remain difficult to predict compared with animal models.
Government and research leaders have welcomed the development, with UK Science Minister Lord Vallance describing it as an example of how artificial intelligence and biomedical research can combine to accelerate vaccine innovation.
