AI-Developed 'Super-Antigen' Vaccine Shows Promise Against Coronaviruses

Artificial intelligence has been used to develop what researchers call a "fundamentally new" type of vaccine designed to protect against broad categories of viruses, potentially preventing future pandemics. A research team at the University of Cambridge says this marks the first occasion where a vaccine's core component was entirely designed by AI and subsequently tested in human trials.

The experimental vaccine targets all coronaviruses, including SARS-CoV-2 variants and animal-borne strains with pandemic potential. While still in early development, the team has begun work on similar AI-designed vaccines for influenza and Ebola. Vaccines typically introduce the immune system to a pathogen’s antigen—the part that triggers an immune response—so it can recognize and fight future infections. However, viruses like influenza and coronaviruses frequently mutate, rendering vaccines outdated and necessitating regular updates, such as annual COVID-19 boosters.

Professor Jonathan Heeney of the University of Cambridge, who led the research, said the traditional approach leaves humanity “always behind” emerging threats. “What we're trying to do is get ahead of the curve,” he stated, “and so far ahead they could protect against new outbreaks or pandemics.” Instead of starting with a live virus strain, the Cambridge team used genetic data from multiple coronaviruses collected by global surveillance programs. An AI system analyzed these genetic codes and designed a synthetic “super-antigen” capable of training the immune system to recognize and respond to an entire viral family—even if the virus mutates or jumps from animals to humans.

Antigens are the active ingredients in vaccines, instructing the immune system what to attack. In this case, the AI-generated antigen was the first of its kind to be tested in humans. Heeney described the technology as “surprising all of us” and emphasized its potential to benefit humanity. “This is about making vaccines that protect us, not just from today's viruses, but protect us from what can cause the next outbreak or disease,” he said. “This is a fundamental shift in how we prepare for pandemics.”

The initial human trial involved 39 participants and focused solely on safety. A second, larger study involving around 200 people is underway to assess immune response. According to findings published in the *Journal of Infection*, the immune response observed was “modest,” though researchers remain encouraged by the results. Professor Saul Faust, who led the clinical trials at the University of Southampton, called the AI design “really exciting” and noted it performed particularly well in scenarios where viruses are evolving rapidly.

The Cambridge team is now advancing similar AI-designed vaccines for seasonal influenza—a universal shot that wouldn’t require annual reformulation—and an H5N1 bird flu vaccine, in response to ongoing outbreaks in poultry. They are also developing a vaccine targeting viral haemorrhagic fevers, including Ebola species. One such outbreak, caused by a previously unvaccinated Ebola strain, is currently affecting communities in the Democratic Republic of Congo.

External experts, including Professor Andy Pollard of the Oxford Vaccine Group, praised the approach as compelling, though he emphasized that real-world human immune systems differ from those in lab animals. Pollard also highlighted AI’s broader potential in vaccine development, calling it a “game changer” that could speed up vaccine design and save lives.

Marian Knight, scientific director at the National Institute for Health and Care Research, called the AI-designed vaccine a “pivotal leap forward,” while Science Minister Lord Vallance hailed it as a “British science success story” that demonstrates the power of combining research expertise with AI to deliver global health benefits.