AI Athlete
Transparency

Data and sources

AI Athlete combines the context you choose with relevant technical and research sources. The origin and purpose of data should remain understandable.

Sources used by the service

Only sources relevant to the feature you use are included in its processing context.

Information you provide

Profile, goals, feedback, injuries, nutrition preferences, race information and other details you choose to enter.

Connected services

Training and activity data from Garmin or Strava when you explicitly connect the corresponding service.

Files you upload

Training files such as FIT, TCX and ZIP, as well as videos submitted for the analysis you request.

Reviewed research sources

Scientific articles and guidelines selected or reviewed for retrieval in relevant AI features.

External services

Depending on the requested feature, the service may use these providers or metadata catalogues.

AI processing

OpenAI generates the requested analysis or recommendation from a minimised, relevant context.

Research metadata

PubMed, OpenAlex, Crossref and Semantic Scholar may help find bibliographic metadata and suitable research candidates.

Data minimisation

Direct identifiers are excluded from AI requests when they are not needed for the requested output.

Your product data

The primary context remains your own training, plan, goals and feedback rather than a generic athlete profile.

How to interpret a recommendationA source list does not mean that every sentence is a direct quotation from a publication. Recommendations can combine your own data, product rules and retrieved research context. Where the product displays specific sources, use them to review the supporting material and apply your own judgement.