Your perspective is yours.

Prism organizes news coverage so you can compare publishers. This policy describes the current website and development app.

What we store

Firebase creates an anonymous reader identifier to deliver the edition and count distinct topics opened each day. Our servers store that daily usage, verified subscription status when applicable, and essential operational records. We don’t need your name or email to browse as a guest.

Your selected perspective and optional source categories are saved locally in your browser or app. They are not a public political profile. Clearing browser data or deleting the app can remove local preferences and anonymous account access.

Reader accounts

You can optionally save a guest account with an email address and password. Firebase Authentication stores account credentials; Prism uses the same reader identity to recognize verified subscriptions across devices. Admin authentication is separate. Verification and password reset emails are sent when requested as part of account setup or recovery.

First-party product analytics

Prism automatically records aggregate service and feature counts to understand reliability and improve the product. These include daily topic-open totals, sessions, publisher-link clicks, Coverage Spotlight and conversation views, paywall interactions and reported failures, split by website, iPhone and iPad. Analytics also include broad landing/page types, observed departures, scroll milestones, foreground active time, session conversions, and bounded traffic categories such as direct, search, social, referral, email and paid. Raw referrer URLs, search terms and arbitrary campaign values are discarded rather than stored. No analytics event includes an article ID, headline, URL, selected perspective, email or free-text content, and we do not use an advertising analytics SDK.

A random web session identifier and bounded acquisition categories are stored in tab-scoped session storage and reused for visits within 30 minutes. A day-specific hash of the authenticated reader ID temporarily supports rate limits and duplicate suppression. A separate pseudonymous account key supports new/returning-reader and day-1, day-7 and day-30 retention counts; it expires after 90 inactive days. This is pseudonymous processing, not a claim that every network request is anonymous. The website respects browser Do Not Track and Global Privacy Control signals. There is no in-app analytics opt-in prompt or toggle.

Temporary deduplication records expire after three days and are removed by a daily cleanup job, normally within four days. Daily aggregate counts are retained for about 400 days. Only the configured, verified administrator account with the explicit analytics role can read the dashboard. Admin sign-in uses a separate browser session and does not replace a reader account.

Approximate country

During an analytics request, the server can estimate a country from the network address using a bundled local DB-IP Country Lite database. Prism analytics records country-level aggregate counts, not raw IP addresses, city, coordinates or GPS location. No location permission or per-reader external geolocation request is used. VPNs, proxies and network routing can affect accuracy. Country data is provided by DB-IP.com under CC BY 4.0.

Notifications

The app requests notification permission on first launch. The reminder preference starts on for new installations, but notifications are enabled only if you grant system permission. You can turn them off in Prism settings or system Settings. If enabled, we store a push token, device identifier, time zone, delivery history and notification preference. These let us send edition summaries during daytime hours. Turn reminders off in Prism settings or revoke permission in system Settings.

Purchases

Apple handles in-app payments. Prism receives signed transaction details to verify the subscription, including product, renewal or expiry information and an account token. We do not receive your payment-card details. Student pricing uses self-declared eligibility; Prism does not collect student documents.

Publishers, X, and AI

Article images, publisher logos and public X avatars or media may load from their providers. Those requests share ordinary network information such as your IP address with the provider. Opening an original-source link takes you to that provider’s own privacy policy.

AI helps match news topics, create short article-grounded feed additions, and discover public X post links during scheduled edition jobs. Posts display original public text rather than AI-written comments. If a lead topic has no publisher image, the feed may use an explicitly labeled abstract AI illustration cached for that shared story. This processing uses public article material; your perspective settings and reading history are not sent with those requests. Article pages retain the original publisher headline.

Retention and control

Preferences remain on your device until reset or cleared. We retain server usage and purchase records to operate reading limits, manage subscriptions and prevent abuse. Notification tokens inactive for over 90 days are excluded from delivery. This version does not yet provide a self-service server-data deletion workflow; clearing local storage does not delete server records.

Service providers

Prism uses Google Firebase for authentication, functions, hosting, data storage and optional messaging, Apple for in-app purchases, and OpenAI for shared editorial processing. We do not sell personal information or use this site for targeted advertising. There are no advertising cookies; first-party analytics use the limited processing described above.

Policy changes

We will update this page as features and data practices change. Review this policy before enabling new optional features.