Use & Processing · Context
WHOOP says its Privacy Principles cover WHOOP Coach and its AI work
WHOOP says its Privacy Principles are the foundation of privacy at WHOOP and apply to everything it does, including its use of artificial intelligence, the creation of machine-learning algorithms and features like WHOOP Coach. This page then walks through how those principles are applied to WHOOP Coach.
Supporting quotations
The WHOOP Privacy Principles are the foundation of privacy at WHOOP. These apply to everything we do, including our use of artificial intelligence, the creation of machine-learning algorithms, and features like WHOOP Coach.
Here is an example of how we apply these Privacy Principles, looking through the lens of the development of WHOOP Coach:
Qualifications & exceptions
- The page describes how WHOOP applies its principles; it is an explanatory article, not a list of contract terms.
- The principles themselves are not reproduced here beyond the five applications described on this page.
Not stated here
- The page does not set out the full text of the WHOOP Privacy Principles or say how they relate to WHOOP's separate privacy policy.
Use & Processing · Company practice
WHOOP says it trains its machine-learning models only on aggregated or de-identified wellness data
WHOOP says that when it trains its machine-learning models it only uses aggregated or de-identified wellness data. It says it removes all identifying information before using member data for that training, so a particular member's personal data cannot be associated with that member. WHOOP says this approach predates WHOOP Coach and has not changed.
Supporting quotations
When we train WHOOP machine learning models, we only use aggregated or de-identified wellness data. WHOOP removes all identifying information before we use any member data for the purpose of training machine learning models.
This approach to training our machine learning models predates the launch of WHOOP Coach and has not changed.
Qualifications & exceptions
- The statement is about training WHOOP's own machine-learning models.
- The page describes de-identification as removing identifying information so data cannot be associated with a member.
Not stated here
- The page does not say what counts as identifying information or how de-identification is checked.
- The page does not say whether members can opt out of this training use.