Abstract
We outline design principles for privacy-preserving modeling on continuous health data, prioritizing on-device processing and data minimization.
Sensitive by nature
Continuous health data is among the most sensitive information a person can generate. Any modeling approach must treat privacy as a first-class design constraint, not a compliance checkbox.
Principles
- Process sensitive signals on-device wherever feasible.
- Minimize what is collected and retained.
- Keep the individual in control of what is shared, and with whom.
Alignment with mission
As a non-profit, our incentives are aligned with patients rather than data monetization — see our Privacy Policy for how these principles are put into practice.
Related: explore the MCT - 1 biosensor and our research overview.