How One Pet Owner Used AI to Design Custom Cancer Treatment

What happened

In 2024, Paul Conningham, a tech entrepreneur from Sydney, faced a dire prognosis for his eight-year-old rescue dog, Rosie, who had aggressive tumors unresponsive to surgery and chemotherapy. Conningham, despite having no medical training, utilized his data science background and AI tools to develop a custom treatment. He first paid the University of New South Wales to sequence DNA from Rosie's healthy and tumor cells to identify specific mutations driving the cancer.

Conningham then employed ChatGPT to analyze this genetic data and create a treatment plan. He also fed Rosie's mutation data into Google DeepMind's AlphaFold AI, which predicted the three-dimensional shapes of her mutated cancer proteins, revealing vulnerable points for immune attack. Armed with this AI-generated targeting information, Conningham convinced university researchers to manufacture a custom mRNA vaccine for Rosie.

The mRNA vaccine taught Rosie's immune system to recognize and attack her unique cancer mutations. The process involved weeks of back and forth with AI-guided simulations and consultations after sequencing, with the first injection in December 2025 following a 2024 diagnosis and other treatments. Rosie received her first shot in December 2025 and a booster in January 2026. Her tumors dramatically shrank, and she is now described as chasing rabbits, a significant improvement over her vet's initial prognosis of months to live. What Conningham did took two months, contrasting sharply with the years and hundreds of thousands of dollars typically associated with human cancer drug trials.

Why it matters

This case signals a fundamental shift in medical discovery and accessibility. Conningham's success demonstrates that AI has reached a point where non-experts can design advanced, personalized medical interventions. This challenges the traditional, centralized model of pharmaceutical development, which typically takes 10 to 15 years and costs billions. The ability for an individual to leverage tools like ChatGPT and AlphaFold to analyze complex genetic data and design a targeted therapy in months, for a fraction of traditional costs, underscores a new era of democratized medicine.

The immediate implications for pet healthcare are significant. Current animal cancer treatments are often expensive, non-specific, and have limited success rates. An AI-driven approach could offer custom vaccines for a fraction of the cost and with fewer side effects, fundamentally changing the math for pet owners. Beyond pets, this model holds potential for human cancer treatment, where personalized therapies can cost hundreds of thousands of dollars and are often inaccessible. AI's ability to slash overall personalized cancer therapy costs by 90% or more could make sophisticated, personalized treatments widely available.

Furthermore, this development reshapes the professional landscape within healthcare. Roles for veterinary technicians and nurses could expand to include AI tool utilization for genetic analysis or personalized medicine coordination. Even small, rural clinics might offer cutting-edge treatments previously limited to major medical centers. The critical insight is that AI handles the intellectual heavy lifting, reducing the need for PhD-level expertise in every step and empowering a broader range of practitioners and determined individuals.

What to watch next

• Observe the pace at which universities and veterinary schools integrate AI-assisted genetic analysis and custom vaccine development into pilot programs.
• Track the commercial availability and cost evolution of personalized pet cancer vaccines, especially from initiatives inspired by Conningham's work.
• Monitor how regulatory bodies adapt to rapidly developed, AI-designed personalized therapies, particularly regarding safety testing and approval pathways for human application.
• Look for evidence of healthcare providers, particularly oncologists, openly embracing and integrating AI tools for patient-specific genetic analysis and treatment planning.
• Assess the growth of online communities and educational resources focused on teaching non-experts to use AI tools for medical data analysis and potential therapeutic design.

What this means for you

Business leaders and operators should recognize that AI is not just optimizing existing processes but enabling entirely new capabilities that can disrupt established industries. In healthcare, this means anticipating a future where specialized knowledge is augmented or even generated by AI, shifting the value proposition from expert analysis to effective application and ethical oversight of AI tools. Explore how AI can empower non-specialists within your organization to tackle complex problems previously requiring extensive training or external consultants.

Begin to understand the practical applications of generative AI and specialized AI models like AlphaFold for data analysis and design in your field. This might involve piloting AI tools for internal data interpretation, product development, or even creating custom solutions to niche challenges. Foster a culture of curiosity and experimentation with these technologies, encouraging your teams to explore how AI can provide "intellectual heavy lifting," allowing more resources to be allocated to implementation, scaling, and customer experience.

Key takeaways

• An individual used ChatGPT and AlphaFold to design a personalized mRNA cancer vaccine for his dog.
• The AI-designed vaccine dramatically shrunk the dog's tumors after traditional treatments failed.
• This case demonstrates AI's potential to democratize personalized medicine, making it accessible to non-experts.
• AI can reduce the time and cost associated with developing targeted cancer treatments significantly.
• The approach could revolutionize personalized cancer care for both pets and humans.

How did a non-expert use AI to treat dog cancer?

A tech entrepreneur, Paul Conningham, without medical training, leveraged AI tools after his dog, Rosie, was diagnosed with aggressive tumors. He sequenced Rosie's DNA to identify specific cancer mutations. Conningham then used ChatGPT to analyze this genetic data and develop a treatment plan. Following this, he fed the mutation data into Google DeepMind's AlphaFold to predict the 3D shapes of Rosie's cancer proteins, which revealed weak points for immune system targeting. This AI-generated information allowed him to design a custom mRNA vaccine, leading to tumor shrinkage.

What AI tools were used for the personalized cancer treatment?

The personalized cancer treatment for Rosie involved two primary AI tools. ChatGPT was used to help analyze the genetic data derived from Rosie's tumor and healthy cells, assisting in the development of a treatment plan. Additionally, Google DeepMind's AlphaFold AI was utilized. AlphaFold predicts the three-dimensional shapes of proteins, and in this case, it showed how Rosie's mutated cancer proteins were folded, which was crucial for identifying targets for the custom mRNA vaccine.

How did the custom mRNA vaccine work?

The custom mRNA vaccine manufactured for Rosie was designed based on AI-generated targeting information derived from her unique cancer mutations. Similar to how some human vaccines work, this mRNA vaccine provided instructions to Rosie's cells. These instructions taught Rosie's immune system to recognize and attack the specific proteins associated with her cancer mutations. This targeted approach allowed her immune system to effectively combat the tumors, leading to their dramatic shrinkage.

What are the broader implications of this AI-driven treatment?

This AI-driven treatment for Rosie's cancer suggests significant implications for the future of personalized medicine. It indicates that non-experts can now design advanced medical solutions, potentially democratizing access to sophisticated treatments. The process, which for Conningham took about two months and incurred a cost significantly lower than traditional drug trials, demonstrates AI's potential to reduce the time and expense of developing targeted therapies. This approach could revolutionize personalized cancer treatment for both pets and humans, making advanced care more affordable and accessible.

What is the potential impact on human cancer treatment costs and accessibility?

The approach used for Rosie's cancer treatment could significantly impact human cancer treatment costs and accessibility. Current personalized cancer therapies can exceed $200,000, making them largely exclusive. AI tools like ChatGPT and AlphaFold can perform the necessary genetic analysis in hours, a task that typically requires months for teams of PhD researchers. By performing the "intellectual heavy lifting," AI could reduce the overall costs of personalized cancer therapies by 90% or more, potentially making sophisticated, personalized cancer vaccines available to a much wider population than is currently possible.