Privacy Policy

NotBroken exists to help you reclaim your freedom, not to profit from your data. That’s why it’s built on privacy by default: we collect only the bare minimum needed for recovery to work, and nothing more.

This policy explains, with no fine print, what each part of the product does (the web, the desktop app, the mobile app and the browser extension), why it does it and what data it touches. The rule is simple: your data lives in YOUR account and on YOUR device; it is never sold, never shared with advertisers and never used to train models.

Account data: your email and a name or alias (which can be anonymous). Used to sign in and to identify you within your own account.

Progress data: your streak and start date, your journal entries, your recovery plan progress, your check-ins, challenges and preferences. All tied only to your account.

Your blocklist: the sites, categories and keywords you choose to block.

Accountability partner (optional): if you set one up, we store your trusted person’s email so we can send them authorization codes. It is not shared with anyone else.

We don’t ask for your real name, phone number, location, or any data that isn’t necessary for the app to work.

Your journal entries, reflections and progress are personal. They’re protected with database-level security (Row Level Security): technically, only your own account can read them.

We don’t read your reflections, we don’t analyze them for commercial purposes and we don’t share them with anyone. The community is anonymous by default: you can post without revealing your identity, and other users never see your email.

What it does: it’s the central hub where you manage your recovery — your streak, journal, plan, community, challenges and blocklist. It’s the “source of truth”: what you set here syncs to the other tools.

What data: it stores in your account only what you enter (described above). We don’t track the pages you visit outside the app and we don’t use third-party ad trackers.

What it does and why: the desktop app applies your blocking at the level of the WHOLE operating system, so it’s effective in any browser and hard to evade in a moment of impulse. To do this, with administrator permission that YOU approve, it can: (1) write your blocklist into the Windows “hosts” file; (2) apply standard Windows policies (the same ones companies use) to force Safe Search, turn off the browser’s “secure DNS” —without which hosts blocking doesn’t work— and optionally block incognito mode; (3) if you turn it on, close browsers that evade blocking (Tor, Opera…) or uninstall Microsoft Store apps you choose (Reddit, TikTok…); (4) install a reinforced task that re-applies the blocking if it’s tampered with.

What data it handles: it signs in with your same account and stores the session locally and encrypted on your own machine (a file in your user folder). It downloads your blocklist from your account to apply it. It does NOT send your browsing history or what you do on the PC to any server: all changes are LOCAL, on your own machine.

Control: all these features are optional and you turn them on. If you set up an accountability partner, turning them off requires their authorization (a code to their email); without a partner, you control them freely. You can remove the blocking and uninstall the app whenever you want.

What it does and why: it blocks sites and filters explicit images inside the browser, mirroring your blocklist.

It works 100% locally: your list of sites and keywords is stored in the browser’s local storage (chrome.storage.local) and never leaves your machine. Blocking and image filtering are processed on your device.

We don’t collect your history: the extension never tells us what sites you visit. It uses no analytics or trackers. The all-sites permission (<all_urls>) is used solely to apply blocking and blur explicit content in your own browser.

It only syncs with your list when you open the app’s website; that’s where it receives the changes you made.

What it does and why: it puts your recovery in your pocket (streak, plan, journal, SOS) and can block adult content on your phone.

Domain blocking uses a LOCAL VPN on the device: it runs on the phone itself and does NOT route your internet traffic through our servers. To filter adult sites it sends DNS lookups (domain names only, never the content of your traffic) to CleanBrowsing’s public family-filter resolver; if it ever fails it “fails open” so your internet keeps working. It does not log the sites you visit.

App blocking (optional): with the Accessibility permission you grant, the app detects which app is in the foreground to block the ones you chose. To do this it reads only the package name of the open app — it does NOT read your screen content, your messages or what you type.

On-screen visual shield (optional, OFF by default): if you turn it on, the app uses Android screen capture to analyze what’s on your screen with on-device AI and instantly cover explicit images. This runs ENTIRELY on your phone: no screenshot or image is ever uploaded, stored or sent anywhere. It uses significant battery, so it’s opt-in and you can turn it off whenever you want.

Maximum Protection (optional, advanced): you can make the app a Device Owner to harden it against tampering (force the filtering DNS, block factory reset and safe boot, grey out the apps you choose). It requires a factory reset to set up and is entirely optional; without it the app works in normal mode.

What data: it connects to your same account and syncs your progress and blocklist. It does not access your contacts, photos or messages.

It’s optional. If you name a trusted person, we store their email for a single purpose: to send them a 6-digit code when you try to turn off a protection or uninstall, so you don’t make that decision alone in a weak moment.

We don’t tell your partner what you do and we don’t give them access to your journal, your activity or your data. They only receive the code you request. Their email is not shared with anyone else and is not used for marketing.

When you talk to the AI Companion, your messages are processed with Anthropic’s API (Claude) for the sole purpose of generating a supportive response in that moment.

These conversations are not used to train AI models and are not shared with third parties for commercial purposes.

Your account data is stored securely on Supabase (encrypted infrastructure). Access is restricted by authentication and row-level security policies, so each user can only access their own.

The desktop and mobile login sessions are stored locally on your device. The email used to send notifications (such as partner codes) is processed through an email provider so it can be delivered to you.

• We don’t sell your data to anyone.

• We don’t show ads or third-party trackers.

• We don’t share your personal information with advertisers.

• We don’t log your browsing history on any device.

• We don’t use your journal or conversations to train models.

• We don’t ask for unnecessary data to function.

You can access, correct or delete your data at any time. From Settings you can export all your data as JSON, and also delete your account.

Deleting your account permanently erases all your associated information: streak, journal, plan progress, posts, blocklist and settings. This action cannot be undone.

On the desktop and mobile apps, uninstalling and removing the blocking leaves your device in its original state.