What we collect, why, and what we never do with it.
Last updated August 6, 2026
DadFM and R.A.D.I.O. are a personal project, not a company with a data-collection roadmap. This policy is written in plain language on purpose — it describes exactly what the app and the engine behind it actually do, not generic boilerplate that doesn't match reality.
DadFM asks for as little as it can get away with, and everything below is either optional or already on your device:
No song file, library export, or bulk upload ever leaves your device. What does get sent is a small amount of text: the one song's title/artist/album/genre, and whatever personalization you've opted into (a nickname, a general area, the current time). That's sent to R.A.D.I.O. only at the moment a line needs to be written or a song needs to be classified — used right then, for that one piece of context, and not stored or reused afterward. There's no standing feed, no background collection, and nothing bundled up and shipped off in bulk.
Podcast episodes work differently, on purpose: since they're something you come back to and listen to again, the topic, any direction you give, the generated transcript, and the generated audio are kept on the proxy server, tied to your account, until it's deleted. We also log basic usage metadata (which endpoint, roughly how much it cost) so the account's usage/cost dashboard works and so a monthly spend limit can be enforced — that's request counts and estimated cost, not the content of what was generated.
Writing and speaking each intro takes two outside AI providers working behind the scenes: one that writes the script from a song's details, and one that turns that script into speech. Each request to them carries only what's needed for that one line — never your library in bulk, never your listening history, never a song file. Per both providers' own published policies, neither trains its models on data submitted through this kind of API use, and neither retains it beyond a short window used only to operate the service and guard against abuse — it isn't kept, reused, or repurposed afterward. We don't control or independently audit their infrastructure, so for the full, current detail, their own policies are the source of truth.
Your library, listening stats, and any personalization details live on your device. Listening stats are encrypted at rest. The bearer secret used to talk to your R.A.D.I.O. proxy, and any nickname/location you've set, are stored in a location explicitly excluded from Android's cloud backup and device-transfer — they stay on that one device, on purpose.
If you create a self-service account, your password is stored as a salted hash — never in plain text, never recoverable, only verifiable. Podcast episodes and basic usage metadata (described above) live on the proxy server for as long as your account exists.
Nickname and location are both opt-in, independently, and can be turned off at any time from Settings — turning location off also clears whatever city-level value was previously resolved. None of this is required for DadFM to work; a completely anonymous listener gets the same station, just without the personal touches.
If you created a self-service account, you're in control of what happens to it — no request form, no waiting on an email reply. Two separate options are available from your account page, once signed in:
Both require re-entering your password, and account deletion requires typing a confirmation phrase — there's no accidental click that erases anything. Neither can be undone once submitted.
If you're in the EU/EEA or UK, this covers your GDPR right to erasure (Article 17) directly. The same self-service pages also cover the right to restrict or stop processing, since deleting your data (or the account) does exactly that. For the right to access or port what's stored about you, your account page already shows your keys, episodes, and usage totals in full — if you want that handed to you as a file instead, or have any other question about your rights, reach out (see "Contact" below) and we'll sort it out directly; this is a small enough project that there's no separate request process to route you through.
DadFM isn't directed at children, and we don't knowingly collect information from anyone under 13.
If what DadFM or R.A.D.I.O. actually collect changes, this page changes with it — no separate notice system exists for a project this size, so the "Last updated" date above is the record of when something last changed.
This is a small, independently-run project. Questions about this policy can be directed to whoever you know that's running it.