Last updated: July 2026. Written in plain language, for a product still in closed beta.
Every message you send passes through a safety check before anything else happens. To generate a response, your message is sent to Anthropic's Claude API, the AI model that powers the conversation. That's the one third party genuinely involved in processing what you write, and it's essential to how the product works, not an advertising or analytics partner.
If a message suggests you may be in crisis, you're redirected toward real human support resources instead of receiving a generated reply. See the note near our waitlist section on the home page for direct crisis resources you can use any time, whether or not you're an Omwerere user.
Account and message data is stored on our own servers (currently hosted on Railway), not sold or shared with data brokers, and not used to train third-party advertising models. We're a small team in closed beta, so infrastructure details may evolve as we grow, and we'll update this page when they do.
If you upgrade to a paid tier, payment is handled directly by Pi Network. We don't store your payment card details ourselves.
Questions about this policy or your data: mwesigyeugene@gmail.com. A real person reads every message.