In a story that’s captivating the AI, biotech, and veterinary worlds, a Sydney-based tech entrepreneur has used artificial intelligence—including ChatGPT and tools like AlphaFold—to help design and deliver a custom mRNA cancer vaccine for his rescue dog, Rosie. The experimental treatment, manufactured with support from scientists at the University of New South Wales (UNSW), has dramatically reduced tumors in the dog diagnosed with terminal mast cell cancer, shrinking a prominent tennis ball-sized mass by roughly 50-75% within weeks to months of administration.
The case, first detailed in The Australian (March 13, 2026) and amplified across social media and tech forums, highlights how accessible AI tools are democratizing complex biomedical workflows—once the domain of large pharma teams and multimillion-dollar budgets.
The Journey: From Desperation to DIY Biotech
Rosie, an eight-year-old Staffordshire–Shar Pei cross rescued from bushland, was diagnosed in 2024 with aggressive mast cell cancer. Traditional treatments—chemotherapy and surgery—failed to halt progression, leaving her with a prognosis of just months to live. Multiple tumors developed, including a large, painful one on her hock (rear leg joint).
Paul Conyngham, a machine learning and data analysis expert with 17 years in AI but no formal biology training, refused to accept the prognosis. He turned to ChatGPT as a brainstorming partner:
- The chatbot suggested exploring immunotherapy and personalized vaccines.
- It guided him to contact the Ramaciotti Centre for Genomics at UNSW for tumor DNA sequencing (cost: approximately AUD $3,000).
- Conyngham uploaded the sequencing data and used ChatGPT to navigate gigabytes of genomic information, identify cancer-driving mutations (notably in the c-KIT protein), and outline steps for an mRNA-based approach.
He further leveraged AlphaFold (Google DeepMind’s protein structure prediction tool) to model how Rosie’s mutated proteins differed from normal versions, pinpointing targets for immune attack.
Conyngham then collaborated with UNSW experts:
- Associate Professor Martin Smith (Director, Ramaciotti Centre) handled the initial “weird” request and genomic analysis.
- Researchers at the UNSW RNA Institute (led by figures like Páll Thordarson) took Conyngham’s AI-generated blueprint and manufactured a bespoke nanoparticle-delivered mRNA vaccine tailored to Rosie’s unique tumor mutations.
Ethics approval (involving extensive paperwork—around 100 pages—and veterinary oversight, reportedly from University of Queensland collaborators) cleared the path. Rosie received her first injection in December 2025.
The Results: Dramatic Tumor Shrinkage and Renewed Vitality
Within about a month of the initial dose (with boosters over the Christmas break and into early 2026), the primary tennis ball-sized tumor on Rosie’s hock shrank by more than half—reports vary from 50% to 75% reduction. Other smaller tumors appeared to recede or stabilize. Rosie regained energy, comfort, and activity levels, dramatically improving her quality of life.
Associate Professor Martin Smith described the outcome as a stunned “holy crap, it worked!” moment, noting the rapid response astounded even seasoned researchers in human oncology.
While not a complete cure (some cancer persists), the shrinkage represents a stunning “first” for a personalized mRNA vaccine in veterinary medicine—especially one guided heavily by consumer-grade AI by a non-specialist.
Why This Matters: From Dog to Human Horizons
Personalized mRNA cancer vaccines—tailored to a patient’s (or animal’s) unique tumor neoantigens—are already advancing in human trials (e.g., Moderna’s and BioNTech’s melanoma/pancreatic programs). But this case stands out for its speed, low cost (~$3k sequencing + AI analysis), and citizen-science element.
Experts see massive implications:
- AI tools like ChatGPT and AlphaFold compressed months of literature review and protein modeling into hours/days.
- It demonstrates how democratized genomics + generative AI could accelerate compassionate-use or off-label experimental therapies.
- For humans, similar approaches (already in phase 2/3 trials) show promise in shrinking tumors when combined with checkpoint inhibitors.
Google DeepMind’s Demis Hassabis and others in the AI community have praised Conyngham’s ingenuity, calling it a glimpse of AI’s role in accelerating precision medicine.
Of course, caveats remain: This is a single case (n=1), veterinary not human, experimental, and not peer-reviewed yet. Regulatory hurdles for human translation are immense—ethics, safety, scalability—but the story fuels optimism that AI could slash costs and timelines for individualized cancer care.
Rosie’s story isn’t just about one brave dog—it’s a proof-of-concept that motivated individuals, armed with AI and university partnerships, can push boundaries in ways once thought impossible.
As one observer put it: “We prepared for AI taking jobs. Turns out it might first help save our pets—and maybe us.”
VFutureMedia will continue tracking developments in AI-driven personalized medicine, from veterinary breakthroughs to human trials.
Ethan Brooks covers the tech that’s reshaping how we move, work, and think — for VFuture Media. He was at CES 2026 in Las Vegas when the world got its first real look at humanoid robots, AI-powered vehicles, and Samsung’s tri-fold phone. He writes about AI, EVs, gadgets, and green tech every week. No hype. No filler. X · Facebook
If you found this useful, the best thing you can do is share it with someone who’d actually appreciate it. And if you want more like it, we’re here every week.
Leave a Comment