KioskNews

KioskNews has no accounts, no advertising, and does not track you between websites. But it is not true that we collect nothing — you can comment here, and comments have to be stored somewhere. This page says exactly what is kept, where, and for how long.

KioskNews is run by one person rather than a company. Questions, complaints, removal requests and reports all reach them directly at support@kiosknews.app.

What we store when you take part

Only two things create a record on our side: posting a comment, and voting on or reporting one.

When you post a comment we store the nickname you typed, the text, the time, and a random browser ID (below). The nickname is not verified, not unique, and not an account — two people can use the same one.

When you like, dislike or report we store which comment, your browser ID, and which action. Nothing else.

The browser ID

The first time you comment or vote, your browser generates a random number and keeps it. It is not built from your device, your screen, your fonts or your network — it is not a fingerprint, and it cannot be used to recognise you on any other website. It exists so that a like can only be undone by whoever placed it, so the same person cannot report one comment fifty times, and to limit flooding.

Clearing your browser data deletes it. After that you are simply a new, unconnected browser to us — and your old comments stay up, because nothing links them back to you.

Your IP address

Every website you visit sees your IP address; that is how a page reaches you. We do not store it alongside your comment. It is used briefly as a counter to stop flooding, in a temporary store that erases itself after five minutes for comments and one hour for translations. Our hosting provider keeps standard server logs under its own retention policy, which we do not control.

What never leaves your browser

Your theme, your saved articles, your nickname, your preferred translation language and your browser ID are stored in your browser only. Two exceptions, stated plainly. First: opening Saved sends the list of saved article IDs to our server so it can look those articles up; that request is answered and not stored or logged. Second: pressing Save leaves an anonymous mark on the article itself — the article’s ID and the time, nothing about you — so our cleanup, which removes old articles after about five days, spares saved ones for sixty days. The mark says “someone, somewhere saved this”; it cannot say who.

Comments are public and lasting

Anything you post is visible to everyone in the world and can be indexed by search engines. Please do not post anything you would not pin to a public noticeboard — including personal details about yourself or anybody else. Comments are kept indefinitely, and an article that has a discussion attached is never deleted.

Reading an article

For most articles we fetch the page and extract a clean reading view on our servers, once, on a schedule. Your browser never contacts the publisher, so reading here does not tell them who you are. When our own fetch is blocked we fall back to r.jina.ai, which fetches the page for us; it sees the article URL and our server’s IP, never yours. Publishers who charge for their journalism are never extracted at all — those link out.

The one exception is pictures: article images load from the publisher’s own servers as you scroll, so those servers do see your IP address, exactly as they would if you visited the publisher directly. They are requested with no referrer, so they are not told which page you are reading. Some publishers put an analytics beacon in the image slot of their feed instead of a photograph — those are detected and discarded, so your browser never requests them.

Translation

This site shows everything in your chosen language: headlines, briefings, articles — and comments, so that people can argue across languages and understand each other. To do that we send the text being translated — an article’s text, a headline, or the words of a comment — to Google’s Gemini API and show you the result. A comment travels as its words only: never with the nickname that wrote it, never with your browser ID or IP address, never linked to who is reading. Every result is stored and reused for everyone who needs that same text in that same language, so most translations do not contact anyone at all, and the original is always one tap away. (If this site is ever run without a translation key, a desktop Chrome browser falls back to translating inside your own machine, sending nothing anywhere.)

The daily briefing

Through the day we group the day’s headlines into stories that several countries are covering, and write a short summary of each. Both steps send published article text to Google’s Gemini API, exactly as translation does. Nothing about you is sent, and the result is the same page for every reader.

Your language

First visit: every browser announces its preferred languages with each request — a standard header, set from your device’s own language setting — and we read it, together with your rough location, to greet you in your language. Neither is stored. If you pick a language in the masthead, that choice is remembered in a single small cookie named kn-read that contains exactly one thing: the two-letter code of your language. It holds no identifier, is sent only to this site, and expires after a year. Deleting it simply returns you to the automatic greeting.

Notifications, if you turn them on

You can ask to be notified when a story you follow moves, or to get the morning briefing at a time you choose. If you do, your browser mints an anonymous delivery address — a one-off URL at your browser maker’s push service — and we store exactly four things with it: that address, its two delivery keys, your chosen language, and your device’s time zone (so nothing wakes you at night — deliveries pause between 22:00 and 07:30 on your clock, and a followed story is mentioned at most once a day). No name, no email, no identifier. Withdrawing permission in your browser kills the address; our next delivery attempt notices and deletes everything we stored. The “turn off” controls on the site do the same immediately.

Asking about a story

Story pages have an “Ask about this story” box. When you use it, your question and published article text are sent to Google’s Gemini API to produce the answer. Your question is not stored by us and is not linked to anything else you do here — but please don’t put personal details in it. Answers come only from articles this site holds, never from the model’s own knowledge.

Who else is involved

• Vercel — hosting, and standard server logs.
• Neon — the database holding articles and comments.
• Upstash — the short-lived counters used for flood protection.
• Google (Gemini) — translation, the daily briefing and story questions, as described above.
• The Movie Database (TMDb) — the trending films and TV in the masthead ticker. The list is fetched by our server; the small poster images load from TMDb’s servers as the page opens, so those servers see your IP, the same as any other picture. This product uses the TMDb API but is not endorsed or certified by TMDb.
• allorigins.win and rss2json.com — fallback feed fetchers, contacted by our server on a schedule and never by your browser.
• Publishers’ image servers — they see your IP when a picture loads.

Cookies, advertising and analytics

No advertising. Readers can be given exactly one cookie, and only by choosing it: kn-read, the two-letter language choice described above — a preference, not an identifier, and never used for tracking. Nothing else sets a cookie for readers; the only other cookie this site can issue is the site administrator’s sign-in, and you will never be given one.

We do count page views, using Vercel Web Analytics. It sets no cookies, stores nothing on your device, and cannot follow you to any other website, which is why you are not asked to consent to it. It tells us how many people opened a page, roughly which country they were in, and which site sent them. It cannot tell us who you are, and it is never joined to anything you write here.

No data is sold, and there is no data broker relationship. There is nothing of yours to sell.

Removing a comment you wrote

There are no accounts here, which means we cannot verify who wrote what — and neither can you. That is a real limitation of an anonymous forum and we would rather state it than pretend otherwise.

In practice: email support@kiosknews.app with a link to the comment and roughly when you posted it. If it plausibly matches, we will remove it. We can also remove everything from one browser ID if you ask and it is clearly yours.

Reporting content

Every comment has a report button, and enough reports hide a comment automatically pending review. For content you believe is illegal rather than merely unpleasant, write to support@kiosknews.app with the link and what you believe it breaches. Those are acted on promptly and by hand.

Children

KioskNews is not aimed at children. Please do not post comments if you are under 13, or under 16 where your country sets that age.

The phone app

The KioskNews phone app is a separate thing and works the way this policy used to describe: no comments, with your sources, saved articles and history kept on your own device. It fetches feeds directly from publishers, falling back to the public proxies listed above and to r.jina.ai for its reading view. Those services see the URL requested and your IP address.

Changes to this policy

If this policy changes we will update the effective date at the top. Continuing to use the site after a change means you accept the update.

Contact

support@kiosknews.app