Microsoft, Mayo Clinic Build Medical AI Model for Patients and Clinicians, With Rollout Taking Years
Updated
Updated · CNN · Jun 2
Microsoft, Mayo Clinic Build Medical AI Model for Patients and Clinicians, With Rollout Taking Years
3 articles · Updated · CNN · Jun 2
Mayo Clinic and Microsoft are developing an AI model trained on medical records, research and clinician expertise, aiming to improve health answers for both doctors and patients.
The push targets a core problem in consumer AI: mainstream chatbots draw on broad internet data, making medical advice unreliable or even dangerous for high-stakes questions.
Mayo Clinic will own the model, first deploying it internally so staff can test accuracy before broader use through hospital tools, patient portals and potentially other health systems.
Microsoft said the effort will take many years to refine for trusted consumer use, though it could eventually also sharpen health responses in Copilot.
The partnership lands as Google, OpenAI and Anthropic expand health assistants, with Mayo betting its anonymized patient data and specialty-care experience can provide an edge.
With AI using your private health data, who is liable when its medical advice goes wrong?
Will a medical 'super-intelligence' fix healthcare or create a new authority that even doctors can't question?
Mayo Clinic and Microsoft Launch Purpose-Built Healthcare AI: A New Standard for Clinical Excellence and Patient Trust
Overview
Microsoft and Mayo Clinic have formed a strategic collaboration to advance healthcare AI, announced at Microsoft’s Build developer event. By combining Mayo Clinic’s clinical expertise and its patient-centric, de-identified data platform with Microsoft’s engineering and AI capabilities, they are developing a frontier AI model tailored for healthcare’s complex needs. This initiative aims to transform healthcare from a traditional pipeline to a platform model, creating innovative solutions for medical applications. The partnership leverages deep clinical context and robust data foundations, ensuring the resulting AI is both powerful and relevant to real-world clinical practice.