AI transparency: the essentials.
- 01
Who provides ONU's AI
ONU's AI features use language models provided by OpenAI, which is listed as a service provider in our privacy policy. AI replies are produced by software, not by a doctor, and are not medical advice. AI features are optional and use your health data only with your consent.
- 02
Know which part of an answer is inferred
A recorded laboratory result, a wearable estimate and an AI explanation have different evidential roles. An explanation should identify the information it relies on and where uncertainty remains. A fluent response is not proof that the underlying conclusion is correct.
- 03
Check numbers against their original source
Names, dates, units, reference ranges and decimal points can be lost during extraction or summarization. Use the original report to resolve differences. If information is missing, an explanation should acknowledge that gap instead of filling it with an assumed result.
- 04
Keep consequential health decisions in care
AI-supported wellness information is not a substitute for diagnosis, prescribing or urgent assessment. Do not change medication or dismiss symptoms because of an automated summary. Bring useful questions and the source documents to a qualified healthcare professional.
- 05
Read the applicable terms and privacy information
The AI terms explain the formal conditions for using the service, while the privacy information addresses handling of personal data. Review those documents for your location and the actual service.
Review your data choices
Learn how ONU handles your information.
