Australian man uses ChatGPT to design AI vaccine for his dog's cancer, catches Sam Altman's eye

When Conventional Treatment Failed, AI Stepped In

Paul Conyngham, a Sydney-based AI consultant, faced every pet owner's nightmare when his eight-year-old rescue dog Rosie was diagnosed with terminal mast cell cancer 1

1

. After nearly a year of misdiagnosis—where veterinarians dismissed visible symptoms as "just a rash"—a biopsy in 2024 confirmed the devastating news 1

1

. With chemotherapy, standard immunotherapy, and surgery proving insufficient as costs mounted, Conyngham refused to accept the grim prognosis. Instead, he embarked on a months-long quest that would demonstrate AI's potential to accelerate medical research in unprecedented ways.

Building a Personalized Experimental Treatment Through AI

Conyngham leveraged AI tools like ChatGPT, Gemini, and Grok to conduct intensive research into emerging cancer treatment options 1

1

. "I would have conversations and just keep them going non-stop" with these chatbots to study cancer therapies in-depth, he explained. Following their guidance, he invested $3,000 to have Rosie's genome sequenced and used the same AI platforms to analyze her DNA data 1

1

. He then turned to AlphaFold, the scientific AI model that won the 2024 chemistry Nobel Prize, to better understand one of the mutated genes driving Rosie's disease 1

1

. This AI-driven initiative focused on developing a personalized mRNA cancer vaccine—technology widely recognized during the Covid pandemic for training the body's immune system to fight specific threats.

From AI Design to Scientific Collaboration

ChatGPT recommended connecting with a University of New South Wales (UNSW) team, which proved instrumental in transforming Conyngham's research into reality 1

1

. Professor Martin Smith, who sequenced Rosie's genome, emphasized that "this was not a clinical trial by any means" and "it's not that AI cured cancer," but rather "it was really driven by his determination to help his dog" 1

1

. Pall Thordarson, director of UNSW's RNA institute which created the vaccine, explained that Conyngham "used the AI program to design the actual mRNA sequence. And then he gave that information to us" 1

1

. The chatbots even assisted in navigating the complex paperwork required for ethical approval, and through his expanding scientific network, Conyngham connected with a University of Queensland professor able to administer the experimental treatment for his dog 1

1

.

Partial Remission and OpenAI CEO Sam Altman's Response

After receiving the custom personalized mRNA cancer vaccine along with powerful immunotherapy in December, Rosie achieved partial remission with her biggest tumor shrinking dramatically 1

1

3

3

. "She regained a lot of mobility and function," Conyngham reported, though not all tumors responded equally well and Rosie has required additional surgery since 1

1

. The story captured global attention when OpenAI CEO Sam Altman called it an "amazing story" in a post on X on Friday, after reportedly meeting Conyngham personally 1

1

2

2

. Altman later noted that "Paul is an extraordinary guy. This should be easy to do, but it is not yet," highlighting both the promise and current limitations of making such innovations accessible 2

2

.

Expert Perspectives on AI's Medical Research Potential

While Conyngham does not claim his findings represent a cure, experts unrelated to the case acknowledge its significance for personalized healthcare. "AI holds lots of promise to improve and accelerate our research strategies," said Nick Semenkovich at the Medical College of Wisconsin 1

1

. However, he cautioned that UNSW and Conyngham "haven't published scientific details outside of their press release and interviews, so we don't know enough about the AI vaccine to understand how much AI helped in its development—or if the vaccine worked the way it was designed" 1

1

. Patrick Tang Ming-kuen from The Chinese University of Hong Kong noted that "AI transforms a 'needle-in-a-haystack' search into a data-driven selection process, drastically shortening the timeframe between diagnosis and vaccine construction," though he acknowledged the real risk of errors 1

1

. Smith emphasized that the combination of "three different disruptive technologies: genome sequencing, artificial intelligence, and RNA therapeutics offers new possibilities and challenges" 1

1

.

What This Means for the Future of Medical Innovation

Since the story went global, Smith's team has been fielding numerous requests: "my cat's got a disease, my dog's got a disease, my aunt has got a disease" 1

1

. But he cautions that "it's hard for us to be able to help" because "there's a lot of things that have to align" 1

1

. Thordarson admits the "short answer is we don't know for sure" what actually led to the reduction in Rosie's largest tumor, and it remains unclear how long she has left to live 1

1

. Experts emphasize that experimental medical interventions, even when supported by AI, require thorough testing, scientific validation, and professional oversight to ensure safety and effectiveness 2

2

. The case demonstrates how AI can enable individuals without formal medical expertise to engage with complex scientific processes, yet the gap between possibility and accessibility remains substantial for widespread application in cancer treatment.