Privacy Policy

Last updated 2026-10-17

Paper Radio turns research papers into audio episodes. This page explains what we keep, why we keep it, and what you can ask us to do about it.

What we keep

• If you create an account: your email address, and the display name and avatar you give us. Signing in with Google shares those with us; a sign-in link uses only your email address.
• Page views: when a page is served, our own web server records the request — the address it came from, the browser's user agent, the page it came from, the path, the status and when — so we can count readers and investigate abuse. That is the access log; it holds no account, and no cross-site identifier beyond the page that linked here.
• What you do here: the episodes you listen to, the ones you like, the shows you follow, the papers you open, and any comments or reactions you post.
• How much is heard: for each play we keep the episode, how many seconds were listened, and where playback stopped, so the station can tell whether an episode is any good. This is counted for every listener, including visitors with no account, and it holds no IP address, no user agent and no cross-site identifier — just the session you already have (if any), an episode and a number of seconds.
• Your studio: the drafts and paper requests you save.
• Fast Track orders: what was ordered and its status. Card details never reach us — you enter them on Stripe's own page.
• Sign-in credentials: if you register a passkey, we store its public key. The private key stays on your device.
• Which day you read: the Morning Preprint remembers, per browser, the last day you read an edition, and whether a play started from one — so the edition can tell whether anybody comes back to it. That is a date on your session and a date on a play, and nothing more: still no IP address and no cross-site identifier.
• What people search for: when a search runs, we keep the words, the tab it was run in and how many results came back — with no account, no IP address and no user agent attached. The search page is not one of them; this is the site's own search, first-party.
• What you type into the ask box: the question is capped and passed to the model provider that writes the answer, with the passage it is answered from; it is not stored in your account, and the station does not train models on it. The access log records that the endpoint was called, not what was asked.

Signing in with Google

If you choose to sign in with Google, Google asks your permission and shares four things with us: an identifier for your Google account, your email address and whether Google has verified it, the display name on the account, and a link to your profile picture. We also keep the identifier against your Paper Radio account, so signing in again later finds the same account rather than making a second one.

We use them to sign you in, to show your name and picture where the site shows you as signed in, and for nothing else. We do not sell them, we do not use them for advertising, we do not use them to train models, and we do not share them with anyone except the people who run the services named above. You can ask us to remove the Google link, or your whole account, by writing to radio@montaan.com.

Paper Radio's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

Cookies and browser storage

• One cookie, pr_session, is what keeps you signed in. It is the only cookie this site sets, and it is also what the Morning Preprint gives a first visit that has none — a guest session, so returning the next day is something the site can tell apart from a visitor arriving for the first time.
• Your sign-in token and your player settings are kept in your browser's own storage, on your device.
• There is no advertising or third-party analytics tracking here, and we do not sell or share personal data for advertising. The play counts described above are our own, first-party, and hold no IP address or cross-site identifier.

Who else is involved

• Google — only if you choose to sign in with Google.
• Stripe — only when you order a Fast Track; it handles the payment, and its own policy covers what it collects.
• Our email provider — when sign-in by email or a podcast-claim mail is enabled, to deliver that message.
• Our model provider — the language model that writes answers, scripts and summaries, and the speech model that speaks them. A question you ask about an episode is sent to it to produce the reply, together with the passage it is drawn from; your account details are not. Whether a hosted provider may use those inputs to improve its own service is governed by that provider's terms, as Stripe's are, and not by us.
• A content delivery network — only if media is served through one; today the site's own host serves it.
• arXiv — the papers themselves: their titles, abstracts, page text and figures come from arXiv, and opening a paper takes you to arXiv's or the DOI's own page.
• Discord — if you follow our community link. It is a separate service with its own terms.

The papers are not ours

Paper Radio summarises work published on arXiv. The papers and their figures belong to their authors and are made available by arXiv under arXiv's own terms; every episode links back to the original and to the DOI.

If you hold rights in a work and want a Paper Radio episode or page that covers it taken down, contact us and we will act on it.

Comments are public; everything else is yours

A comment you post is visible to everyone and is shown with your display name. You can delete your own comments.

Your listening history, your likes and your follows are kept for you. We publish counts — how many people liked an episode, how many played it — never who.

How long we keep it, and what you can ask for

Page views in the access log are deleted after 180 days. Everything else is kept while your account exists, or while the station needs it to work: plays and searches as counts, your drafts and follows until you remove them, comments until you delete them, and Fast Track orders for as long as tax and payment records require.

Ask us and we will delete your account and the data that belongs to it, give you a copy of what we hold in a machine-readable file, correct what is wrong, or restrict what we do with it; you may also object to processing we base on our legitimate interest, and ask for your data portability. Write to radio@montaan.com and we answer within 30 days. If you are not satisfied, you may complain to your data-protection authority.

We process account and order data to perform our contract with you, access-log and play data on the basis of our legitimate interest in keeping the station secure and knowing what is listened to, and anything optional only with your consent.

Children

Paper Radio is not directed at children. If you are below the age at which you can consent to this kind of service where you live, please do not create an account.

Changes to this policy

When this policy changes we update the date below. If a change is significant we will say so on the site.

Last updated: 27 September 2026.