ChatGPT did not cure a dog’s cancer

When an Australian tech entrepreneur with no background in biology or medicine said ChatGPT helped save his dog from cancer, the story spread with the kind of validation Big Tech had long craved: proof that AI would revolutionize medicine and tackle one of its deadliest diseases. The reality, as usual, is more complicated.

The version of the story that made the rounds online, first reported by The Australianwas relatively simple. In 2024, Sydney-based Paul Conyngham was diagnosed with cancer. Chemotherapy slowed the disease but failed to shrink the tumors. After vets said “nothing could be done” for the Staffordshire bull terrier-shar pei, Conyngham said: “I took it upon myself to find a cure.”

Conyngham said he used ChatGPT to brainstorm treatment ideas. The chatbot presented immunotherapy as an option and referred him to experts at the University of New South Wales, who then established the genetic profile of Rosie’s cancer. He then used ChatGPT and Google’s protein structure AI model, AlphaFold, to help make sense of the results. With the help of UNSW’s Professor Pall Thordarson, he developed a personalized mRNA vaccine tailored to Rosie’s tumor mutations. Thordarson said The Australian he believes this is the first time such a treatment has been designed for a dog.

Within weeks of Rosie’s first injection last December, Conyngham said her tumors had shrunk and she was doing better, even chasing rabbits in the park. However, they did not entirely disappear and one tumor did not respond at all. “I have no illusions that this is a cure, but I believe this treatment has gained Rosie much more time and quality of life,” Conyngham said. The Australian.

This nuance was lost as the story spread. News week headlined “Owner with no medical training invents cure for dog’s terminal cancer,” while the New York Post said a “tech pro saves his dying dog by using ChatGPT to code a personalized cancer vaccine.” On social media, many accounts touted Rosie’s case as a “cure” and a sign that a new era of personalized medicine had arrived. Some, including OpenAI Chairman and Co-Founder Greg Brockman, certainly should have known, and others, like Google DeepMind CEO Demis Hassabis, did and shared it without hype. Elon Musk also joined in, keen to point out that xAI’s Grok also played a role – a detail that was missing from much of the original coverage.

History also gives far too much credit to AI. Not only was Rosie not cured of cancer, but it is not clear whether the mRNA vaccine was responsible for her improvement. The personalized treatment was given alongside another form of immunotherapy known as a checkpoint inhibitor, designed to help the immune system target tumors, making it difficult to know whether the vaccine had any effect. One of the scientists involved, Martin Smith, said the team was carrying out tests to check the immune response.

The vaccine itself was also not generated by a chatbot. ChatGPT did not design or create Rosie’s treatment; human researchers have done it. At most, the chatbot served as a research assistant helping Conyngham analyze the medical literature – impressive, but far from the breakthrough involved.

Reports are also vague about AlphaFold’s role. David Ascher, professor and director of biotechnology programs at the University of Queensland in Australia, said: The edge that the model “could contribute to structural hypotheses about proteins, but it is not a turnkey system for cancer vaccine design.” The official guidelines, he noted, also caution that AlphaFold is not validated to predict the effects of certain mutations nor does it model “multiple biologically important contexts.”

Grok’s contribution is even more difficult to pin down. On Ascher said Grok would realistically fall into the same category as ChatGPT: a tool that “could help search literature, summarize articles, translate jargon, suggest workflows, write code or documents, and help a user think through options.” A useful role, but far from what the design of a cancer vaccine suggests.

Overall, Ascher said Rosie’s case “is best viewed as evidence of unusual and very specific possibility rather than a pattern that ordinary people can easily replicate.” It required “substantial” expert work, he said, “not just a chatbot and a few prompts.”

This distinction is particularly important in medicine, where success depends not only on the production of plausible information but also on the expert physical labor of producing, testing, and administering an actual treatment. Alvin Chan, assistant professor at Nanyang Technological University in Singapore, who is developing AI for biomedical and pharmaceutical discoveries, said: The edge the “AI created this” framing ignores this massive human effort, without which “the AI ​​output would have remained simple text on a screen.” In Rosie’s case, AI is better understood as a tool for sketching out a plan than as the creator of the treatment itself.

The whole thing gives off a slight smell of a publicity stunt that is difficult to shake. Bold claims built on dubious foundations and using vague methods fit perfectly into the world of tech fundraising. mRNA vaccines – much like the broader promise of personalized medicine – remain largely unproven as cancer treatments in humans, let alone dogs, and while the case may be real, it seems too simple and conveniently glosses over the tens of thousands of dollars and significant expertise required to turn the idea into a viable treatment.

I contacted Conyngham to request a discussion on X, but have yet to receive a response. His profile reads “Ending Cancer in Dogs” and links to a Google form describing his “dream of making this process something everyone could have access to.” The form asks if your dog has cancer, if you are a researcher or scientist looking to get involved, and if you are an investor.

I think it would be a mistake to dismiss Rosie’s story as completely meaningless. AI may not replace laboratories anytime soon, but it makes science more accessible to ordinary people. However, this does not mean making care more accessible, and few patients – or pet owners – have easy access to the world-class experts, specialized equipment and substantial funds needed to turn this information into real treatment.