The story about a dog named Rosie and a custom cancer treatment quickly spread online with claims that ChatGPT helped cure her disease, but the facts tell a different story and show a far more limited role for AI in this case.
Rosie, a Staffordshire bull terrier-shar pei from Sydney, was diagnosed with cancer in 2024, and while chemotherapy slowed the disease, it did not shrink the tumors, which pushed her owner Paul Conyngham to search for alternative options and explore new ideas using AI tools.
According to The Australian, Conyngham used ChatGPT to explore possible treatments and came across immunotherapy, which led him to researchers at the University of New South Wales, where experts analyzed Rosie’s cancer and helped create a personalized mRNA vaccine based on her tumor mutations.
The treatment did show some improvement, as Conyngham said, “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,” and while some tumors shrank after the first injection, others did not respond and the cancer did not disappear.
At the same time, doctors gave Rosie another immunotherapy treatment called a checkpoint inhibitor, which makes it difficult to know whether the vaccine played a clear role in her progress, and researchers are still testing how her immune system reacted.
AI’s role was limited
Experts have pushed back on claims that AI designed or created the treatment, since human researchers handled the actual science, testing, and delivery, while ChatGPT mainly helped with reading research and understanding complex data.
David Ascher, a professor at the University of Queensland, said AlphaFold “could contribute structural hypotheses about proteins, but it is not a turnkey cancer-vaccine design system,” which highlights that these tools assist research but do not replace it.
Alvin Chan from Nanyang Technological University explained, “the AI’s output would have remained just text on a screen,” which shows that real medical work depends on experts, lab testing, and resources.
The case shows how AI can support research and help people understand complex medical information, but it also shows the risks of overstating what these tools can do, especially in areas like cancer treatment where results depend on careful testing and expert work.
Rosie’s case stands as a unique example that required significant expertise and funding, and it does not offer a simple path that others can easily follow or repeat.
