What happened
Researchers at the University of Cambridge, working with DIOSynVax, used AI and machine learning to design a "super-antigen" entirely inside computer simulations, the first vaccine ingredient built this way to be tested in people. In a Phase 1 trial of 39 healthy volunteers published in the Journal of Infection, the needle-free DNA vaccine, delivered by a micro fluid jet, proved safe and aimed not just at SARS-CoV-2 but at a whole family of related coronaviruses, including the original SARS and the bat viruses that could spark the next pandemic.
Why this matters: most vaccines chase the last outbreak. Professor Jonathan Heeney compares the usual approach to a dog chasing its tail. Designing a single antigen to cover an entire viral family, pan-Sarbecovirus, flips the model from reactive to pre-emptive, and the same method could extend to flu and Ebola.
A safe Phase 1 in 39 people is the start, not the finish, and Phase 2 is still ahead. But "AI-designed, broad-spectrum, needle-free" is a genuinely different starting point for pandemic preparedness. If we can design vaccines against viruses that have not emerged yet, how should that change the way we fund and stockpile them?
Source
Reported by AI-designed universal coronavirus vaccine passes first human trial via sciencedaily.com, published June 5, 2026.