COUNTY KERRY, IRELAND — A 37-year-old Irish father’s reliance on ChatGPT for medical advice has resulted in a delayed diagnosis of stage-four esophageal cancer, highlighting growing concerns about the inappropriate use of artificial intelligence tools for healthcare decisions.
Warren Tierney, a former psychologist from County Kerry, sought advice from OpenAI’s ChatGPT chatbot earlier this year when he began experiencing persistent sore throat and swallowing difficulties. The AI tool repeatedly reassured him that cancer was “highly unlikely” and described his symptoms as showing “no red-flag symptoms,” leading Tierney to delay seeking professional medical care for several months.
Background: Turning to AI Instead of Doctors
Tierney’s case reflects a troubling trend of individuals turning to AI chatbots for medical guidance. Over several weeks, he described his symptoms to ChatGPT, which consistently provided reassuring responses. In one interaction documented in the case, the chatbot empathetically pledged to “walk with you through every result that comes” whether cancer was present or not, even offering to draft legal affidavits or buy him a Guinness if its advice proved wrong.
The AI’s confident, conversational tone and statistical reassurances gave Tierney what he later described as a false sense of security. He acknowledged that he “maybe relied on it too much,” trusting the chatbot’s assessment over his own concerns about worsening symptoms.
The Misdiagnosis and Its Consequences
The delay in seeking professional medical care proved catastrophic. When Tierney’s symptoms eventually worsened to the point where he required emergency hospital treatment, doctors diagnosed him with stage-four esophageal adenocarcinoma—a rare and aggressive form of throat cancer with a five-year survival rate between five and ten percent globally.
The two-month delay in diagnosis meant that what might have been treatable cancer at an earlier stage had progressed to its most advanced form. Esophageal adenocarcinoma is notoriously difficult to treat once it reaches stage four, with limited treatment options and poor prognosis outcomes.
Tierney’s family has since launched a European fundraising campaign to cover experimental treatments abroad, as conventional treatment options have become severely limited due to the advanced stage of the disease.
Medical Expert Reactions
Healthcare professionals have expressed alarm at Tierney’s case, emphasizing the critical risks of relying on AI for medical diagnosis. Medical experts stress that persistent symptoms, particularly those lasting beyond two weeks and involving swallowing difficulties, voice changes, or unexplained weight loss, require immediate professional evaluation.
“Early medical evaluation including physical examination, imaging, and possibly biopsy is essential to differentiate benign causes from serious conditions such as throat or esophageal cancers,” medical professionals warned in response to the case.
The medical community has consistently advised that AI tools should only serve as general informational guides, never as substitutes for professional diagnosis or treatment plans. Experts emphasize that AI models lack clinical experience, cannot process physical examinations, and are not continuously updated with the latest medical research.
Recent research has highlighted AI’s limitations in medical settings. Studies show that while AI models demonstrate promise in certain screening applications, they still lack the sensitivity and specificity required for clinical use and may produce biased or incomplete assessments.
Wider Context: AI’s Growing Role in Healthcare
Tierney’s case occurs amid a broader debate about artificial intelligence’s appropriate role in healthcare. While AI tools offer unprecedented access to medical information, they also present significant risks when used inappropriately.
Recent studies have revealed a concerning trend: AI companies have dramatically reduced medical safety warnings in their outputs. Research indicates that fewer than 1% of AI-generated responses to medical questions included warnings in 2025, down from over 26% in 2022.
Despite these risks, AI continues to advance in legitimate medical applications. Researchers are developing AI tools that can diagnose certain cancers, guide treatment protocols, and predict patient survival rates with increasing accuracy. However, these applications involve trained medical professionals using specialized AI systems—not consumers consulting general-purpose chatbots.
The healthcare AI market faces ongoing challenges regarding regulation, ethical oversight, and public education about appropriate use. Medical experts advocate for clear guidelines distinguishing between AI as a research tool for professionals and its inappropriate use for self-diagnosis by patients.
Current Status and Lessons Learned
Tierney’s current condition remains serious as he battles stage-four cancer while his family seeks experimental treatment options through international medical centers. His case has become a cautionary tale about the dangers of substituting AI advice for professional medical care.
The incident underscores several critical lessons: the importance of seeking immediate medical attention for persistent symptoms, the limitations of AI in providing personalized medical judgment, and the need for continued public education about appropriate AI use in healthcare contexts.
OpenAI has reiterated that ChatGPT is not designed to provide medical advice or treatment, emphasizing its role strictly as an informational aid. The company continues to stress that users should always consult qualified healthcare providers for medical concerns.
As AI technology becomes increasingly sophisticated and accessible, Tierney’s story serves as a stark reminder that while these tools offer valuable information resources, they cannot replace the expertise, clinical judgment, and diagnostic capabilities of trained medical professionals. The case highlights the urgent need for clear public guidelines on AI use in healthcare and the potentially life-threatening consequences of overreliance on artificial intelligence for medical decisions.
