Is Rosie the dog's AI-designed mRNA cancer vaccine legitimate?
In 2025, Sydney tech entrepreneur Paul Conyngham used ChatGPT and AI tools to help design a personalised mRNA cancer vaccine for his rescue dog Rosie, a staffy–Shar Pei cross diagnosed with mast cell cancer. After conventional chemotherapy slowed but failed to shrink the tumours, Conyngham used AI to help sequence Rosie's tumour DNA, identify mutations, and design an mRNA vaccine. The vaccine was manufactured at the UNSW RNA Institute and administered under ethics approval at the University of Queensland. Rosie received her first injection in December 2025, followed by a booster in early 2026. Results became visible about a month after treatment, with the primary tumour shrinking by approximately 75%, though at least one tumour did not respond. This resolves based on my subjective judgment of whether the result appears to be a legitimate, real therapeutic effect — not a fluke, misattribution, or hype cycle. I will resolve by end of 2027. Examples of things that would push toward YES: -Rosie shows sustained tumour reduction or long-term survival well beyond what vets expected -The involved scientists (Thordarson, Allavena, Smith) publish or present the case formally -Other dogs are treated with a similar approach and show comparable responses -A peer-reviewed study validates the methodology -Credible oncologists or immunologists publicly endorse the approach as scientifically sound Examples of things that would push toward NO: -Rosie's tumours return quickly and her outcome is roughly what was expected without the vaccine -Scientists involved distance themselves from the story or clarify that results were overstated -Experts identify that the tumour shrinkage is better explained by delayed chemotherapy effects or spontaneous regression -The story quietly disappears with no follow-up data, replication, or scientific publication - Evidence emerges that key claims (e.g. the 75% shrinkage figure) were exaggerated or poorly measured Feel free to clarify in comments. Update 2026-03-15 (PST) (AI summary of creator comment): The role of AI in the vaccine design is not the primary resolution criterion — the market resolves based on whether the therapeutic effect is real. However, the creator notes that if AI's contribution was trivial (e.g., just finding a phone number), that would weigh negatively, while meaningful AI involvement (e.g., writing code to identify mutations) would count positively. The creator is less skeptical that an LLM did most of the digital steps, and more skeptical that the result is novel and effective.
