ChatGPT didn't cure dog cancer, but AI tools helped researchers develop personalized treatment

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 couldn't help but spread. It's the kind of validation Big Tech has long craved: proof that AI will revolutionize medicine and take on 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 Australian, was relatively straightforward. In 2024, Sydney-based Paul Conyngham learned that his dog Rosie had 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 mix, Conyngham said "I took it upon myself to find a cure." Conyngham said he used ChatGPT to brainstorm treatment ideas. The chatbot surfaced immunotherapy as an option and pointed him toward experts at the University of New South Wales, who then genetically profiled 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 professor Pall Thordarson, he pursued a personalized mRNA vaccine tailored to Rosie's tumor mutations. Thordarson told The Australian he thinks it's the first time such a treatment has been designed for a dog. A few weeks after Rosie's first injection last December, Conyngham said her tumors had shrunk and she's doing better, even chasing rabbits in the park. They've not disappeared entirely, though, and one tumor didn't respond at all. "I'm under no illusion that this is a cure, but I do believe this ­treatment has bought Rosie ­significantly more time and quality of life," Conyngham told The Australian. That nuance was lost as the story proliferated. Newsweek ran the headline "Owner With No Medical Background Invents Cure for Dog's Terminal Cancer," while the New York Post declared that a "Tech pro saves his dying dog by using ChatGPT to code a custom cancer vaccine." On social media, many accounts hyped Rosie's case as a "cure" and a sign a new era of personalized medicine had arrived. Some, notably OpenAI president and cofounder Greg Brockman, should have definitely known better, while others, like Google DeepMind CEO Demis Hassabis, did and shared it without hype. Elon Musk joined in too, keen to point out that xAI's Grok also played a part -- a detail that was absent from much of the original coverage. The story also gives AI far too much credit. Not only was Rosie not cured of cancer, it's not clear the mRNA vaccine was responsible for her improvement. The personalized treatment was administered alongside another form of immunotherapy known as a checkpoint inhibitor, designed to help the immune system target tumors, making it difficult to know if the vaccine had any effect at all. One of the scientists involved, Martin Smith, said the team is performing tests to check the immune response. Nor was the vaccine itself generated by a chatbot. ChatGPT did not design or create Rosie's treatment; human researchers did. At most, the chatbot served as a research assistant helping Conyngham parse medical literature -- impressive, but a far cry from the breakthrough implied. Reports are also vague on AlphaFold's role. David Ascher, a professor and director of biotechnology programs at the University of Queensland in Australia, told The Verge that the model "could contribute structural hypotheses about proteins, but it is not a turnkey cancer-vaccine design system." Official guidance, he noted, also warns that AlphaFold is not validated for predicting the effects of some mutations and does not model "several biologically important contexts" either. Grok's contribution is even harder to pin down. On X, Conyngham wrote that "the final vaccine construct for Rosie was designed by Grok," but it's not clear what that means in practice or what inputs the model was given. Ascher said Grok would realistically fall into much the same category as ChatGPT: a tool that "could help with literature search, summarising papers, translating jargon, suggesting workflows, drafting code or documents, and helping a user think through options." A useful role, but hardly what "designing a cancer vaccine" suggests. All in all, Ascher said Rosie's case "is better seen as an unusual, highly specific proof of possibility than a template ordinary people can readily reproduce." It needed "substantial" expert labor, he said, "not just a chatbot and a few prompts." That distinction is especially important in medicine, where success depends not just on producing plausible information, but on the expert, physical work of producing, testing, and delivering actual treatment. Alvin Chan, an assistant professor at the Nanyang Technological University in Singapore who is building AI for biomedical and drug discoveries, told The Verge the "AI made this" framing ignores this massive human effort, without which the "AI's output would have remained just text on a screen." In Rosie's case, AI is better understood as a tool for sketching a blueprint than as the creator of the treatment itself. The whole thing carries a faint whiff of a PR stunt that is hard to shake. Bold claims built from questionable foundations using vague methods comfortably fit inside 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 feels too tidy and conveniently glosses over the tens of thousands of dollars and significant expertise required to turn the idea into a viable treatment. I reached out to Conyngham asking for a chat on X but have not received a response. His profile says "Ending Cancer for Dogs" and links to a Google form describing his "dream to make this process something everyone could have access to." The form asks whether your dog has cancer, whether you're a researcher or scientist who wants to get involved, and whether you are an investor. I think it would be a mistake to dismiss Rosie's story as completely meaningless. AI may not be replacing the lab anytime soon, but it is making science more accessible to ordinary people. However, that's not the same as making care more accessible, and few patients -- or pet owners -- have ready access to the world-class experts, specialized equipment, and substantial funds needed to turn that information into real treatment.

A man used AI to help make a cancer vaccine for his dog - an oncologist urges caution

An Australian tech entrepreneur has helped create what appears to be a made-to-measure cancer vaccine for his dog, Rosie, using artificial intelligence tools such as ChatGPT as part of the process. The science behind this sounds intimidating - DNA sequencing, mRNA vaccines, "neoantigens" - but at its core, it is about reading the instructions inside a tumour and then writing a new set of instructions to help the immune system see it. Rosie is an eight-year-old rescue Staffordshire bull terrier cross that developed aggressive mast cell cancer, a common skin cancer in dogs. She had surgery and chemotherapy, but the disease kept coming back and she ended up with large, ugly tumours on her leg. Vets told her owner, Paul Conyngham, that she probably had only months to live. Instead of accepting that, he decided to use the tools he knew from his day job in tech - data analysis, AI and coding - and apply them to his dog's cancer. Decoding the tumour The first step was to understand what made Rosie's tumour different from her healthy cells. Every cell in the body carries DNA - a long, chemical molecule that acts like a biological instruction manual. You can think of DNA as a very long string of letters written in a four-letter alphabet. Cancer happens when enough of those letters change, by chance or through damage, so that some cells start to grow and divide out of control. Sequencing a tumour's or normal cell's DNA is essentially reading through that long string of letters and comparing it to the "normal" version to see where it has gone wrong. A lot of my own research has focused on this. Conyngham paid a university lab to sequence the DNA from Rosie's tumour. That produced a huge file listing the mutations - the spelling mistakes in the cancer's instruction manual - that set her tumour apart from her healthy tissues. On their own, those files are just data. The question is what to do with them. This is where he turned to an AI chatbot. He asked it how scientists design personalised cancer vaccines and how he might go from a list of mutations to specific targets for a vaccine for Rosie. A cancer vaccine in this context is different from the childhood vaccines we are used to. Traditional vaccines prevent infections: you give someone a harmless version or fragment of a virus or bacterium so their immune system can "learn" to recognise it in advance. A cancer vaccine, by contrast, is usually therapeutic rather than preventive. It is given to someone who already has cancer, with the aim of training their immune system to spot markers on the cancer cells that it has previously ignored and then attack them. This is where mRNA comes in. If DNA is the master instruction book, mRNA (messenger RNA) is more like a photocopied page that gets sent to the cell's protein-making machinery - think of it as a short piece of code that carries a single command: "make this protein". Some of the COVID vaccines use mRNA: they deliver a strand of mRNA that tells our cells to make the spike protein from the coronavirus, so the immune system can practise on it. The body then breaks down the mRNA; it does not change your DNA. For a personalised cancer vaccine, scientists choose small parts of proteins that are unique to a particular tumour - so-called neoantigens - and encode them in an mRNA sequence. When this mRNA is injected, cells take it up and briefly make those tumour-linked protein fragments. The immune system can then see these fragments and, ideally, begins to treat any cell displaying them as abnormal and dangerous. In effect, it is using mRNA to give the immune system a "most wanted" poster for that individual cancer. With help from AI tools, Conyngham sifted through Rosie's tumour mutations to pick out candidates that might make good neoantigens. He also used protein structure prediction software to model how some of these mutated proteins would look, trying to guess which ones would be visible to her immune system. Crucially, he did not manufacture a vaccine in his garage. Once he had a shortlist of targets, he approached researchers at the University of New South Wales, including experts in RNA technology, who reviewed the data and designed an mRNA construct based on it. Their team turned this digital design into a physical mRNA vaccine in the lab. It was a one-off product, made just for Rosie, encoding several of the mutations in her tumour. She then received this experimental vaccine at a veterinary research centre, with booster doses over the following months. Reports from her vets and owner suggest that several tumours shrank markedly, her overall tumour burden fell, and her energy and behaviour improved. One resistant tumour has prompted a second round of analysis and a follow-on vaccine targeting a different set of mutations. Promising, but not a cure It should be noted that this is a single dog, not a controlled study, and mast cell tumours can behave unpredictably. We cannot be sure how much of Rosie's improvement is due to the vaccine, how long it will last, or whether the same approach would help other dogs, let alone humans. The AI did not "cure cancer" by itself. It acted as an always-available guide and assistant, but qualified scientists still had to check its work and do the hard parts in the lab. Even so, this case is a vivid example of several ideas coming together. DNA sequencing allows you to read the specific mutations in an individual cancer. mRNA technology lets you quickly write a custom set of instructions to show those mutations to the immune system. AI systems make the complex biology more navigable for non-experts, suggesting possible targets and explaining concepts - though their outputs still require expert scrutiny. Put those together, and something that would once have required a major pharmaceutical programme - a bespoke cancer vaccine - can now be attempted, at least experimentally, for a single animal. For the informed public, perhaps the most important point is not that AI has magically solved cancer, but that the basic ingredients of high-end personalised medicine are becoming more accessible. A motivated dog owner can now order tumour DNA sequencing, ask an AI to help interpret it, and partner with an academic lab to turn that interpretation into an mRNA vaccine. A significant scientific and ethical challenge ahead is to develop methods for testing such approaches properly, protect patients and animals from false hope and unsafe experiments, and determine who should have access if they prove to be effective.

An Australian tech entrepreneur used AI to help create the first-ever bespoke cancer vaccine for a dog to treat his beloved pet Rosie | Fortune

In 2024, Sydney tech entrepreneur Paul Conyngham found out his dog Rosie had cancer. But after attacking the diagnosis with chemotherapy and surgery, the tumors persisted and Rosie got sicker. So he turned to AI and eventually developed a custom a mRNA cancer vaccine with the help of Australian scientists. Most of Rosie's tumors have shrunk, and the dog is back chasing rabbits. OpenAI's ChatGPT suggested immunotherapy and directed Conyngham to the University of New South Wales Ramaciotti Center for Genomics, according to a report in the Australian. While Conyngham doesn't have a background in medicine, he is an electrical and computing engineer who cofounded Core Intelligence Technologies. He was also a director for the Data Science and AI Association of Australia. After reaching out to university, he convinced researchers there to help him and paid UNSW for Rosie's genomic sequencing. Then he started digging into the DNA. "I went to ChatGPT and came up with a plan on how to do this," Conyngham told the Australian. He also used AlphaFold, an AI tool from Google's DeepMind, to find mutated proteins that could be potential targets for treatment. While an immunotherapy treatment that looked like a good fit for Rosie was identified, the drugmaker wouldn't provide it. Then nanomedicine medicine pioneer Pall Thordarson, director of UNSW's RNA Institute, stepped in and used Conyngham's data to develop a bespoke mRNA vaccine in less than two months. "This is the first time a personalized cancer vaccine has been designed for a dog," he told the Australian. "This is still at the frontier of where cancer immunotherapeutics are -- and ultimately, we're going to use this for helping humans. What Rosie is teaching us is that personalized medicine can be very effective, and done in a time-sensitive manner, with mRNA technology." Rosie got her first injection of the cancer treatment this past December, then received a booster in February. Most of her tumors have already shrunk dramatically. And while they haven't disappeared, Rosie's health has improved. In a thread on X Saturday, Thordarson said Rosie's story demonstrates that technology can "democratize" the process of designing cancer vaccines. He cautioned that Rosie may not be cured as some tumors haven't responded to the vaccine, though it bought her more time. Still, Conyngham will take it. "In December she had low energy because the tumors were creating a huge burden for her," he told the Australian. "Six weeks post-treatment, I was at the dog park when she spotted a rabbit and jumped the fence to chase it. I'm under no illusion that this is a cure, but I do believe this ­treatment has bought Rosie significantly more time and quality of life." Rosie's journey has stunned some people in the tech world while also pointing to AI's potential to produce breakthroughs in medicine, perhaps turning diagnoses once considered death sentences into routine ailments. Matt Shumer, cofounder and CEO of OthersideAI, took to X over the weekend to flag a story about Conyngham and his dog. "This is what I mean when I say the world is going to get very weird, very soon," he wrote. "Expect more stories like this, each sounding increasingly more insane."

Did ChatGPT Really Cure a Dog's Cancer? It's Complicated - Decrypt

AI tools assisted with research and data exploration, but did not design the cancer therapy, despite headlines saying so. OpenAI co-founder Greg Brockman amplified a widely shared story over the weekend about a dog treated with a personalized mRNA cancer vaccine developed with help from ChatGPT, drawing attention across tech and AI communities. The case centres on Rosie, a seven-year-old Shar Pei owned by Australian AI consultant Paul Conyngham. According to posts circulating online, Rosie had been given only months to live before receiving the experimental treatment, which Conyngham said was developed with assistance from the AI chatbot. "Back in 2022, I noticed strange lumps on her head," Conyngham wrote in a November 2024 thread documenting the journey from the beginning. "What the vet deemed as 'just warts' ended up being late-stage cancer." Vets estimated Rosie had between one and six months left and told Conyngham there was nothing more they could do. The account spread rapidly after Brockman shared it with his hundreds of thousands of followers, prompting coverage across several technology outlets. While the treatment itself appears genuine, the role ChatGPT is credited with in developing the vaccine has been debated, with some researchers questioning how much of the process could realistically be handled by a large language model. Conyngham said he didn't give up on Rosie. Instead, he decided to build a research pipeline out of consumer AI tools. He started with ChatGPT, using it to design a plan of attack. The model told him he needed genomic sequencing, one sample of healthy tissue and one from the tumor, and pointed him toward specific institutions and equipment. "The most ironic thing is that in a previous chat session with ChatGPT, it said that I should attempt to reach out to Elita or Dr. Martin and that I should use an Illumina machine," he wrote at the time. So he followed that lead. A director at UNSW connected him to Dr. Martin Smith, head of the Ramaciotti Centre for Genomics, who agreed to sequence Rosie's genome for around $3,000. Ten days. Thirty-times depth in healthy tissue, 60-times in tumor: the higher pass rate needed to isolate the mutations driving the cancer. The Centre returned 320 gigabytes of raw data. Genomic information is expressed in strings of the letters A, T, C, and G, so experts essentially ended up with a stack of 700,000 double-sided pages full of only those four letters, the University of New South Wales reported in June of last year. That was Rosie's genome, her biological fingerprint. He then focused on c-KIT, a protein well-documented in the published literature on mast cell tumors in dogs. Using Google's AlphaFold, he modeled Rosie's version of the protein and compared it against the healthy baseline. It looked wrong, mutated in ways that matched what the literature predicted. He then searched for existing compounds that might attack c-KIT or proteins similar to it, and found one: a drug already in use in the U.S. to treat a different cancer in humans. "We took her tumour, sequenced the DNA, we converted it from tissue to data, and we used that to find the problem in her DNA and then develop a cure based on that," Conyngham told the Australian Today Show on Saturday. "ChatGPT assisted throughout that entire process." Even so, there's a big gap between ChatGPT finding a cure for cancer and ChatGPT assisting in research. Conyngham eventually connected with Prof. Palli Thordarson, Director of the UNSW RNA Institute. "Prof. @martinalexsmith performed the DNA/RNA sequencing to convert Rosie's tissue into raw data," Conyngham posted. "Prof @PalliThordarson assembled the mRNA vaccine," he added in another tweet. Thordarson confirmed this in his own thread: "Proud with @UNSWRNA to have been involved & making the mRNA-LNP for Rosie," he wrote on X on Sunday. "The intersection of RNA technology, genomic & AI poses an opportunity to change the way we do medicine and make access more equitable." But Dr. Smith wasn't a man behind a ChatGPT screen. He was a professor running a university RNA institute, doing what his lab was built to do. And when Conyngham identified the final vaccine construct -- the specific molecular blueprint that would be encoded into the mRNA -- he revealed which tool designed it. Not AlphaFold. Not ChatGPT. "The final vaccine construct for Rose was designed by Grok." That said, he recognized in a separate post that "Gemini did a ton of the heavy lifting too." ChatGPT was used to sift through scientific papers and identify researchers who might be able to help. The chatbot pointed to the Ramaciotti Centre and suggested sequencing equipment suited to the task, functioning largely as a tool for navigating the research literature. That role can be useful, but it differs from designing a vaccine or performing scientific analysis. AlphaFold, a deep-learning system from Google DeepMind, predicts three-dimensional protein structures from amino acid sequences. It's not the first model trained on biological data: other open-source initiatives like Ankh or AlphaGenome work on similar premises. Conyngham used Alphafold to model Rosie's c-KIT protein. The rendering carried a confidence score of 54.55, which UNSW structural biologist Dr. Kate Michie publicly described as low. She noted that AlphaFold "can get stuff wrong" and that significant lab work is needed to validate any output. Dr. Smith, the UNSW genomics director, confirmed publicly in the same thread that AlphaFold was not, in fact, used for the mRNA vaccine design at all. Dr. Thordarson was careful about the framing, too. "This may not have cured Rosie," he wrote on X. "Bought time for sure, yes, but some of the tumours didn't respond." His team is now checking whether those tumors mutated differently, which would explain why parts of the treatment worked and others did not. The vaccine also did not work in isolation. "The treatment required co-administration of a checkpoint inhibitor," Thordarson noted, "likely to be with all personalized cancer vaccines." The use of AI for cancer treatment has not always been a history of success. In 2017, internal IBM documents revealed that Watson for Oncology, marketed as a system that could recommend cancer treatments better than human oncologists, was generating what its own engineers flagged as "unsafe and incorrect" recommendations. MD Anderson Cancer Center abandoned the project after spending $62 million on it. IBM sold off Watson Health in its entirety in 2022. The Rosie case doesn't fall into the category of AI failures. No one was harmed, the underlying science is established, and the researchers involved have recognized credentials. The mRNA platform itself is supported by clinical research. The unease lies more in how the story has been framed. When AI tools receive credit for work carried out by scientists and research institutions, it can blur public understanding of what the technology actually does. The researchers who performed the sequencing, produced the vaccine, and managed the safety protocols risk fading into the background. The episode offers a reminder that AI can assist with tasks such as navigating scientific literature, but it remains far from replacing the expertise and infrastructure required to design and produce medical treatments.

A DIY Medical Miracle? How One Man Used ChatGPT to Help Create a Custom Cancer Vaccine for His Dog

When tech entrepreneur Paul Conyngham found out that his dog Rosie had deadly mast-cell cancer, he was devastated. When chemotherapy and surgery didn't work to shrink the tumors, the electrical and computing engineer who co-founded Core Intelligence Technologies, took matters into his own hands and turned to AI for help. After giving OpenAI's ChatGPT Rosie's health information, it suggested immunotherapy and specifically recommended that he contact the doctors and researchers at the University of New South Wales Ramaciotti Center for Genomics. "The first step was to reach out to the university to get Rosie's DNA sequenced. The idea is you take the healthy DNA out of her blood and then you take the DNA out of her tumour and you sequence both of them to see exactly where the mutations have occurred. It's like having the ­original engine of your car and then a version of the engine 300,000km down the road -- you can compare them and see where there's damage," he explained to the Australia's Today. When associate professor Martin Smith saw the work that Conyngham had already done, he was amazed. "He called and told me he had ­analyzed the data and found ­mutations of interest and then used AlphaFold (an AI program) to find the proteins that were ­mutated, and then identified ­potential targets and matched them to drugs, and he was ­wondering could I help him find someone to synthesise this compound that he'd identified. I'm like, 'Woah, that's crazy!' I was motivated by his enthusiasm,'' he told the Australian.

Watching his dog slowly die, techie refused to give up. Then he used AI and created a custom 'cancer vaccine' for his pet friend

Australian tech expert Paul Conyngham used artificial intelligence to help save his dog Rosie, a Staffy-Shar Pei mix, after she was diagnosed with advanced cancer. When conventional treatments like chemotherapy failed, he sequenced her DNA, identified mutations causing the tumour, and worked with the University of New South Wales to develop a personalised mRNA vaccine. Rosie received her first injection in December, followed by booster doses, and her tumour has since shrunk, with her energy and quality of life improving.

Engineer crafts custom dog cancer vaccine with ChatGPT; Rosie's tumors shrink 75%

An experimental, personalized mRNA cancer vaccine designed for a Sydney dog named Rosie has shown rapid tumor reduction after classical treatments failed. Within weeks of administration, Rosie's tumors were reported to have shrunk by approximately 50% to 75%. Her condition improved markedly, with increased activity and a return to behavior observed before the illness. The scale and speed of the change surprised both the owner and senior researchers. Veterinarians gave up Paul Conyngham, a tech entrepreneur and A expert, pursued the project after veterinarians gave Rosie only a few months to live following a mast cell tumor diagnosis in 2024. He sequenced DNA from both tumor and healthy cells to identify mutational changes and isolate neoantigens -- defective proteins found only on cancer cells. He then used a neural network to analyze results and prioritize targets with algorithms, including Google DeepMind's AlphaFold, to model protein structures. AI tools, including ChatGPT, helped him plan the work and decipher complex genetic data, streamlining a discovery process that in traditional settings can take months or years. https://x.com/pubity/status/2033214798159122745?s=20 Conyngham independently developed a treatment strategy and formulated a vaccine tailored to Rosie's tumor profile. He relied on expert collaboration to translate his computational findings into a therapeutic product. Scientists at the University of New South Wales (UNSW) and the University of Queensland assisted in synthesis. The UNSW RNA Institute's Professor Páll Thordarson completed the vaccine in under two months. A first for a dog Professor Thordarson confirmed it is "the first time a personalised cancer vaccine has been designed for a dog," according to International Business Times UK. He said the approach holds potential for human applications, while emphasizing the need for rigorous studies. Associate Professor Martin Smith of UNSW described the initial request to sequence a private individual's dog as unusual, noting Conyngham's persistence. After sequencing, AI methods filtered thousands of observed mutations to find those most likely to elicit an immune response. Protein-structure predictions then helped prioritize targets for a bespoke mRNA sequence. The outcome has encouraged researchers to consider whether similar pipelines could accelerate development of patient-specific vaccines against other cancers in people. "AI allows us to do things that were previously impossible for a single person," Conyngham said.