Designed with activists' security as our priority
Effective April 19, 2026 · This policy covers the web service at outcryai.com. For the iOS and macOS app, which runs entirely on-device, see our App Privacy Policy.
Privacy by Design: Outcry AI was built with activists' security in mind. We collect no personal identifying information and implement privacy-preserving techniques throughout our system. Your data remains yours alone—we scrub identifying information from all requests, and offer one-click complete data deletion. Guest users can engage without creating accounts, and all communications use secure, anonymized identifiers. Our code is structured to protect your privacy by default, ensuring your activism remains confidential and resistant to surveillance.
We don't collect or store personal identifying information. Guest accounts use only timestamps as identifiers, and we don't track IP addresses or device information.
Our middleware automatically strips identifying headers and uses privacy-preserving fingerprinting techniques instead of raw identifiers for essential operations.
A one-click "Delete my data" option permanently removes all your chat history and content. For registered users, this also deletes your account.
Outcry AI reserves the right to use conversations with free-tier and guest users — past and future — to train, fine-tune, and evaluate future Outcry AI models. This is how we keep Outcry free and improving for activists who cannot pay. Before your conversations enter a training dataset we strip IP addresses and hash identifiers.
Your conversations stay inside Outcry. We do not publish user conversations, and we do not sell them to third parties. Real user text is used for internal training only. If we release training datasets publicly — for other activist AI projects to build on — those datasets contain only synthetically generated conversations produced by our models, not text written by users.
Your conversations as a paid subscriber are never used for training unless you explicitly opt in. Opt-in lives in Settings · Privacy and is reversible at any time.
Your conversations may be used to train future Outcry models internally. They are not published or sold. You can delete your entire history to pull it back before training, or upgrade to paid to opt out going forward.
Guest sessions are ephemeral and are used for internal training. Create an account (free or paid) to keep your conversations linked to you and to get a privacy choice.
If we release training data publicly for other activist AI projects, those datasets contain only synthetically generated conversations produced by our models — never real user text. Per-conversation consent is logged in the database so internal training exports are auditable against each user’s choice.
Conversations with the on-device Outcry AI never leave your device and are never used for training. A separate, app-specific privacy policy governs the iOS/macOS app. Read the App Privacy Policy →
While we've designed Outcry AI with strong privacy protections, please remember that no digital platform can guarantee complete security against determined adversaries with significant resources. For highly sensitive activism in high-risk contexts, consider using additional security measures like Tor Browser.
Try Outcry AI with confidence, knowing your privacy is our priority. No account required to start.