TechCrunchAI & LLMs

Amazon's Health AI Assistant Goes Mainstream, Raising Familiar Privacy Questions

Amazon has moved its healthcare-focused artificial intelligence, Health AI, from its exclusive perch within the One Medical service to its main website and shopping app. The tool, a product of Amazon's $3.9 billion acquisition in 2023, is now available to any U.S. user, regardless of Prime membership.

The assistant performs a range of functions, from answering general wellness questions to managing prescription renewals and booking appointments. Its more advanced features, however, require users to connect their personal medical records via the national Health Information Exchange. With that access, it can interpret lab results and offer guidance tailored to an individual's history.

This expansion places Amazon in direct competition with other AI giants like OpenAI and Anthropic, which have recently launched their own specialized health chatbots. Amazon is offering an incentive: Prime members get up to five free text consultations with a One Medical clinician for common conditions.

The company's announcement immediately renewed discussions about data privacy in sensitive AI applications. Amazon states that interactions are HIPAA-compliant, protected by encryption, and that its models are trained on abstracted patterns, not directly identifiable data. Yet, the company did not detail its encryption methods or specify which employees might review conversations—a common practice for improving AI systems. TechCrunch has inquired for those specifics.

For now, users who sign up on Amazon's health page will be notified when access is granted. They can then pose questions directly within the familiar Amazon interface, testing whether a retail giant can become a trusted health confidant.

Source: TechCrunch

← Back to News