A man with no biology degree used ChatGPT to help build his dog a cancer vaccine. It worked.

A man with no biology degree used ChatGPT to help build his dog a cancer vaccine. It worked. First, the headline going around the internet, “Man Cures Dog” is wrong. Nobody cured a dog. The word cure is doing a lot of heavy lifting in a story that is genuinely remarkable without it, and veterinary professionals deserve the accurate version.

Here is what actually happened.

Paul Conyngham is a machine learning engineer in Sydney with 17 years of experience in data science and AI. He is not a biologist. He is not a veterinarian. In 2024, his eight-year-old rescue dog Rosie — a Staffordshire Bull Terrier-Shar Pei cross — was diagnosed with mast cell cancer, the most common skin cancer in dogs and one that is typically incurable through conventional treatment. Conyngham tried surgery, chemotherapy, and immunotherapy. Chemo slowed the spread but nothing shrank the growths. Vets gave her one to six months. 

So he did what a data scientist would do. He went to ChatGPT and started asking questions.

What he actually built and how

ChatGPT functioned as a research assistant. It helped Conyngham navigate unfamiliar biomedical literature and plan his analysis steps.It pointed him toward the University of New South Wales Ramaciotti Centre for Genomics, where he paid $3,000 to have Rosie's tumor DNA sequenced. He then used AlphaFold, Google DeepMind's protein structure prediction tool, to model the mutations driving her cancer and identify potential targets.

The output was a half-page mRNA sequence formula. Páll Thordarson, director of the UNSW RNA Institute, took that formula and produced the physical vaccine in under two months. Rachel Allavena at the University of Queensland, who already held ethical approvals for experimental canine immunotherapies, administered it. Conyngham drove ten hours for Rosie's first injection in December 2025 and returned for boosters.

The tumor on Rosie's leg shrank by 75% within a month of her first injection. Scientists at UNSW and the University of Queensland confirmed the result. They also confirmed this is a single dog, a single tumor, and precisely zero controlled trials. 

One tumor responded. Another did not. Conyngham is already working on a second vaccine targeting the resistant growth.

What this is and what it is not

The framing bouncing around social media, that "AI designed a cancer vaccine," is doing a lot of heavy lifting. Awesome Agents The AI accelerated a pipeline that would otherwise have taken months of manual literature review. It did not replace the immunologists, the RNA engineers, or the veterinary oncologists who made the vaccine real. Conyngham's 17 years of machine learning experience mattered enormously. This was not a layperson with a chatbot. This was a technically sophisticated person using AI tools to bridge into an adjacent field, supported at every critical step by credentialed scientists who donated their time and expertise.

"I'm under no illusion that this is a cure," Conyngham told The Australian. That framing is worth repeating to every client who comes in having read the social media version.

Why the veterinary profession should pay attention

Thordarson said Rosie's story demonstrates that technology can "democratize" the process of designing cancer vaccines. That is the real headline. Not that a dog was cured, she was not, but that a personalized mRNA cancer vaccine was designed, produced, and administered for a canine patient in under two months, at a fraction of the cost that would be involved in a human clinical trial. Moderna and Merck's version of a personalized mRNA cancer vaccine, when approved for humans, is expected to cost somewhere between $100,000 and $300,000 per patient.

Personalized cancer vaccine research for humans is already well advanced. Merck and Moderna's mRNA-4157 vaccine combined with Keytruda showed a 49% reduction in melanoma recurrence or death at five years in a Phase IIb trial. Phase III trials are underway for melanoma and non-small-cell lung cancer. The science Conyngham accessed for Rosie is the same science that is moving through human oncology pipelines right now.

Mast cell cancer is the most common skin tumor in dogs. It is sitting in your waiting room. The idea that personalized immunotherapy could eventually become part of the standard of care for canine oncology is not science fiction. Rosie's case is one dog, no controlled trial, and a set of circumstances that required a very specific combination of scientific access, technical expertise, and institutional goodwill that most dog owners will never have.

But the direction of travel is clear. The tools are getting cheaper, faster, and more accessible. The veterinary profession has an opportunity to be ahead of this conversation rather than reactive to it — and that starts with understanding exactly what happened in that lab in Gatton, Australia, and exactly what it does and does not mean.

Rosie recently jumped a fence to chase a rabbit. Her owner considers it the most meaningful data point of the entire project.