AlphaMED Resources · https://alphamedicine.uk/Resources/cases/ai-vaccine

AI designed coronavirus vaccine trial

Scientists at Cambridge used machine learning to design one vaccine meant to protect against a whole family of coronaviruses. A first trial in 39 healthy volunteers found it safe. Cambridge said in June 2026 that the technology could prevent pandemics, though the trial was never meant to show whether the vaccine works.

Updated 12 October 2026 (4 minute read)

AutonomyBeneficenceNon maleficenceJustice
  1. Professor Jonathan Heeney founds the company DIOSynVax
  2. First volunteers are given the vaccine
  3. Last of the 39 volunteers is vaccinated
  4. Results are published in a medical journal
  5. Cambridge announces the results

Summary

  • A team led by Professor Jonathan Heeney at the University of Cambridge used machine learning to design the antigen for a new vaccine.
  • The antigen is the part of a vaccine that the immune system learns to attack.
  • Covid and flu vaccines copy one strain of a virus, so they have to be updated as the virus changes.
  • The team built this one from features shared by a whole family of coronaviruses. The family includes the viruses behind SARS and Covid, and bat viruses that could jump to people.
  • Cambridge announced on 5 June 2026 that the technology "could prevent future pandemics before they begin".
  • The evidence so far is one small safety trial. Scientists outside the team say it has not yet shown that the vaccine protects anyone.
  • Researchers gave the vaccine to 39 healthy adults between December 2021 and September 2023, in NHS hospitals in Southampton and Cambridge.
  • It was a phase 1 trial, the first test in people. Phase 1 looks for side effects and the right dose, and larger trials come afterwards.
  • The paper reported no significant safety concerns. It calls the immune response "modest", in volunteers who already had strong immunity to Covid.
  • Cambridge says the vaccine needs more development before public use. A second trial of around 200 people comes next.
Addenbrooke's Hospital

Dig deeper

  • Waves of the Omicron variant of Covid passed through during the trial, which made the immune results hard to read.
  • The paper's own summary describes the vaccine as "computationally designed". The phrase AI designed comes from the university's announcement.
  • An AI scientist at the drug company Novo Nordisk, who was not involved, said the trial did what phase 1 is meant to do.
  • He added that it "did not yet show the strong, broad immune response you would want".
  • Professor Andy Pollard of the Oxford Vaccine Group told the BBC the animal evidence was compelling and the real test is in people.
  • Professor Saul Faust, who led the trial, said the design "definitely has potential".
  • Heeney founded DIOSynVax in 2017, the Cambridge company that developed the vaccine with the university.
  • That company, the University of Regensburg and Cambridge's commercial arm own the rights to the vaccine.
  • Innovate UK, the government's innovation agency, paid for most of the trial.
  • The trial's entry in the public register of trials said the side effects were "currently unknown", with most expected to be mild. It listed "no known benefits to participants".
  • A 2014 study in the BMJ, a medical journal, found that when UK university press releases exaggerated a health finding, most news stories based on them did too.

Ethical issues

PrincipleHow it applies here
AutonomyThe volunteers were healthy adults, whom the law presumes to have capacity, and consent to a trial counts only if it is "given freely" after the person is told its "nature, significance, implications and risks". Here the side effects were unknown and the hospital offered payment for volunteers' time.
BeneficenceThe regulations in force during the trial said a trial should go ahead "only if the anticipated benefits justify the risks", counting the benefit to "other present and future patients". The register listed no known benefit to the volunteers, so the benefit weighed here belongs to people outside the trial.
Non maleficenceNobody had been given this vaccine before, so the first volunteers had 0.2 mg and later groups had up to 1.2 mg. The same regulations said the safety of trial subjects "shall prevail over the interests of science and society", which pulls against beneficence when the hoped for benefit is to everyone else.
JusticeA government agency paid for most of the trial and NHS hospitals hosted it, and the rights belong to the company, another university and Cambridge's commercial arm. The paper says the needle free form was used to support future use in places with few resources, so fairness here covers who shares the reward and who gets the vaccine.

Interview questions

  • What do you know about the AI designed coronavirus vaccine tested at Cambridge?
  • What can a phase 1 trial tell us, and what can it not?
  • The trial's register listed no known benefit to volunteers and said the side effects were unknown. Why might someone still take part, and is it right to pay them?
  • A patient tells you they have read that AI has made a vaccine that will stop the next pandemic. What do you say?
  • What are the arguments for and against announcing early results like these to the public?
  • The lead scientist founded the company that developed the vaccine. How should researchers handle an interest like that?
  • Public money paid for most of this trial. Who should own a vaccine that comes out of it?
  • Why might some people be wary of a vaccine designed by AI, and how would you respond to them?

Sources

  1. cam.ac.ukcam.ac.uk/…/new-universal-vaccine-technology-could-protect-us-from-futur
  2. medicalnewstoday.commedicalnewstoday.com/…/world-first-vaccine-designed-by-ai-tested-humans-will-it-liv
  3. pubmed.ncbi.nlm.nih.govpubmed.ncbi.nlm.nih.gov/…/42155675
  4. NHSnhs.uk/…/clinical-trials
  5. bbc.co.ukbbc.co.uk/…/crrpggegwe0o
  6. enterprise.cam.ac.ukenterprise.cam.ac.uk
  7. GOV.UKgov.uk/…/innovate-uk
  8. isrctn.comisrctn.com/…/ISRCTN87813400
  9. pmc.ncbi.nlm.nih.govpmc.ncbi.nlm.nih.gov/…/PMC4262123
  10. legislation.gov.uklegislation.gov.uk/…/1
  11. legislation.gov.uklegislation.gov.uk/…/2022-01-01
  12. cuh.nhs.ukcuh.nhs.uk/…/pioneering-needle-free-covid-vaccine-begins-trials

Photo: Rob Johnson, CC BY-SA 2.0, via Wikimedia Commons, colour altered.

Practise answering these with a mentor

Our interview preparation has you answer questions like these out loud, with feedback on each answer.

See interview preparation Book a free consultation

alphamedicine.uk/Services · alphamedicine.uk/Contact

Related

ADHD and autism overdiagnosis debate AutonomyBeneficenceNon maleficenceJustice Far more people in England are diagnosed with ADHD and autism than 25 years ago. People disagree over whether ordinary difficulties are being given medical labels, or people who were missed for years are at last being recognised. A government review reported on 9 October 2026 that it could not show overdiagnosis on a large scale, partly because the data are out of date. It wants help to be given on need, without waiting for a diagnosis. Corridor care in NHS hospitals AutonomyBeneficenceNon maleficenceJustice Corridor care means looking after patients in corridors, waiting rooms and other parts of a hospital never built for it, because every bed is full. The government and NHS England both call it unacceptable. NHS England still counted about 3,300 patients a day in September 2026. Vaping and the generational smoking ban AutonomyBeneficenceNon maleficenceJustice Fewer than one in ten UK adults now smoke, and more people in Great Britain vape than smoke. Vapes help smokers quit, but about 1.1 million people who never smoked use them too. Under a law passed in April 2026, nobody born in 2009 or later will be sold tobacco from January 2027.

More NHS hot topics, medical ethics case studies and medical school interview questions are in AlphaMED Resources, which we add to and update through the application cycle.