The news: Amazon One Medical rolled out an AI health assistant for its primary care members. The Health AI tool, powered by LLMs on Amazon Bedrock, lives in the One Medical app and is intended to support, not replace, a doctor’s visit.
How it works: Health AI is designed to help patients answer health questions, book appointments, and manage medications based on their medical data.
Core features include:
- Explaining lab results and medical images
- Offering guidance on symptoms, progress, conditions, and possible treatments based on users’ medical records, flagging when a doctor’s input is needed
- Guiding users to the appropriate care option based on their needs
- Scheduling visits with a One Medical provider
- Submitting prescription renewal requests
Why it matters: Consumers are increasingly turning to AI tools for health information, chiefly because they provide fast and easy-to-understand answers to their medical questions. Receiving quick responses was the top reason (43%) US adults use AI tools for health queries, per a June 2025 Drip Hydration survey.
One Medical’s Health AI launch follows recent health tool rollouts from two major AI players.
- Earlier this month, OpenAI announced ChatGPT Health, which allows users to receive health guidance after uploading medical records and data from some wellness apps.
- Last week, Anthropic released Claude for Healthcare, extending similar features to Pro and Max subscribers.
However, One Medical says Health AI differs by understanding a patient’s health context without requiring uploads of medical records from multiple health systems (assuming that One Medical already has the information). Also, One Medical is bound by HIPAA to protect medical data—unlike most consumer health apps. This could ease concerns about data privacy, though some prospective members might remain wary given Amazon’s ownership of the primary care company and the potential use of health data for advertising.
Implications for healthcare AI players and providers: General-purpose AI companies like OpenAI and Anthropic that are entering healthcare are counting on large user bases to trust them with health data, but they lack a care delivery component to pair with AI guidance. Following the ChatGPT and Claude launches, concerns emerged about whether always-available AI health tools could steer users away from doctor visits by answering questions using their medical histories.
Now, we’re seeing healthcare providers like One Medical responding by embedding AI assistants into clinical care, with the added advantage of connecting users to a doctor. Other tech-focused medical providers will likely follow with patient-facing AI assistants, positioning AI-driven guidance as an extension of existing care rather than a substitute. Success for both provider AI assistants and general-purpose chatbots will hinge on strong medical data protections and AI guidance that is tightly grounded in patients’ documented histories.