For the first time in history, a vaccine component designed entirely by artificial intelligence has been tested in humans. And it worked.
Researchers at the University of Cambridge, working alongside biotech firm DIOSynVax Ltd, announced on June 5 that their AI-engineered vaccine candidate completed a Phase I clinical trial with no significant side effects. The trial involved 39 healthy volunteers between the ages of 18 and 50, conducted at Addenbrooke’s Hospital and the University of Southampton.
The vaccine doesn’t just target one virus. It’s designed to protect against the entire Sarbeco group of coronaviruses, which includes SARS-CoV-2 (the virus behind COVID-19), the original SARS virus, and a range of bat coronaviruses that could potentially jump to humans.
How AI built a better antigen
The research team fed genetic sequence data from a broad range of coronaviruses, collected through global surveillance efforts, into AI and machine learning systems. The AI then identified the viral features that remain consistent, or “conserved,” across multiple strains and species.
From those conserved features, the AI synthesized what the researchers call a “super-antigen.” This is a single engineered protein designed to train the human immune system to recognize the structural elements that coronaviruses simply cannot mutate away without losing their ability to function.
Professor Jonathan Heeney, the lead researcher on the project, described the approach as a shift from reactive vaccine development toward a “future-proof” model. The preliminary results published in the Journal of Infection showed that the vaccine triggered promising immune responses against multiple target viruses, not just one specific strain.
What the trial actually showed
Phase I trials are designed to answer one question above all else: is this thing safe? On that front, the Cambridge vaccine passed cleanly. None of the 39 participants experienced significant side effects.
The researchers observed immune responses targeting multiple coronaviruses across the Sarbeco group. Phase II and Phase III trials will be required to prove efficacy at scale.
The researchers have also suggested this same AI-driven methodology could be applied to other viral threats, including influenza and Ebola.