Research & Product Improvement Consent

Your optional permission to use de-identified insights from your conversations to improve menthra's AI models and publish aggregate wellness research.

Product improvement that uses personal content is fundamentally different from providing the service. It deserves its own explicit consent, completely separate from your core account. You do not have to give this consent to use menthra — the service works exactly the same way whether you opt in or not.

• We can use de-identified data derived from your conversations (patterns, themes, outcomes) to improve AI model behavior.
• We can include de-identified insights in published aggregate research and transparency reports — for example, reporting on what types of emotional patterns menthra helps users work through most effectively, or how different companion interaction styles correlate with user-reported progress over time. No individual is ever identifiable in these reports.
• No direct identifiers are ever included. We follow the HIPAA Safe Harbor standard (45 CFR §164.514(b)) at minimum.
• We will never sell your data. This consent does not change that.

The service works exactly the same. Your data is not used for model training or research. We still use aggregate, de-identified operational data (like load metrics) to keep the platform running — that is covered by legitimate interest, not this consent.

Opt out at any time in account settings. Withdrawal stops future research use; it does not retroactively remove your contribution from already-published aggregate statistics (we cannot un-bake a loaf of bread). For data deletion, see Your Rights.