AI Chatbots Provide Inaccurate Recommendations for Blood Pressure Monitors

Thumbnail Image

The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.

A study presented at the American Heart Association's Hypertension Scientific Sessions 2026 found that leading AI chatbots, including ChatGPT, Copilot, Perplexity AI, and Google Gemini, frequently misidentified clinically validated home blood pressure monitors, with accuracy rates ranging from 63% to 83%. This poses risks of health mismanagement.[AI generated]

Why's our monitor labelling this an incident or hazard?

The AI systems involved are explicitly mentioned as AI-powered search tools (Google Gemini, ChatGPT, Copilot, Perplexity) used to recommend medical devices. Their inaccurate outputs have a plausible direct link to harm to health (harm category a) because users relying on these recommendations may use unvalidated devices, resulting in incorrect blood pressure readings and subsequent misdiagnosis or mistreatment. Since the harm is realized or highly likely due to the AI's inaccurate recommendations, this qualifies as an AI Incident rather than a hazard or complementary information.[AI generated]
AI principles
SafetyRobustness & digital security

Industries
Healthcare, drugs, and biotechnology

Affected stakeholders
Consumers

Harm types
Physical (injury)

Business function:
Citizen/customer service

AI system task:
Interaction support/chatbotsOrganisation/recommenders


Articles about this incident or hazard