What this site indexes, and what it never touches

World Wires is a news monitor. It reads public feeds, public APIs, public web pages and a small curated set of Tor onion services, stores headlines, short summaries, links and metadata, and shows them with their source, trust tier, origin network and verification state. Every item links back to where it came from.

Dark-net material

Only curated lists: onion mirrors of established outlets, index pages of ransomware leak sites named by a public tracker, release listings, a site-status directory. Nothing is discovered by crawling the Tor network.
Text and metadata only. No files, archives, torrents or images are downloaded, stored or mirrored.
Items from these sources are labelled unverified claims by criminal groups unless an established outlet or authority reports the same event. They are shown so the public can see what is being claimed, not as fact.
Captcha- or login-gated sites are never accessed, and nothing is taken from directories that carry abuse material.

Article text

For items from established outlets on the clear net, the article text is fetched once, honouring robots.txt and the TDM reservation signal, to index it for search, to answer questions from it, and to extract names and places. At most a short excerpt is shown, always with a link to the publisher; the full text is never republished, is kept for 14 days, and is removed on request. No PDFs, images or media are fetched.

Satellite view

For some stories the site shows public satellite images of the reported place: Copernicus Sentinel-2 (contains modified Copernicus Sentinel data, via Microsoft Planetary Computer), NASA GIBS daily imagery and NASA FIRMS fire detections. The images are fetched and stored by this server, so opening a story sends no request to those services. Flags such as visible change or fire detected are computed from the images and fire data without AI. A short machine reading of the images may be shown; it is written by an AI model, is labelled as such, and is never a verification of the story. No press or news photographs are shown.

Research careers

Job postings, PhD positions, fellowships and funding calls are listed as title, employer, place, closing date and a short summary, with a link to the original posting where the application is made. They are kept out of search engine indexes and removed from this site once they close (after 60 days when no closing date is given).

State media and official accounts

Some channels are state media, government-aligned, or official government accounts. They are indexed because what they claim is itself news, and every item from them carries a badge (state-aligned, official account) and never counts as corroboration on its own. Outlets listed under the EU broadcasting ban (Council Regulation (EU) 2022/350 and later sanctions packages) are excluded: the source catalogue refuses them. Sanctions data comes from the UN Security Council list and from OpenSanctions (CC BY-NC 4.0). Conflict events on the map may come from ACLED (Data: ACLED, research licence, non-commercial); market data points from FRED (St. Louis Fed).

AI-generated text

Translations, one-line summaries, topic labels and the rolling briefing are produced by a language model from the indexed sources only. They can be wrong. The original headline is always available.

Removal requests

If you are named in an indexed item and believe it is unlawful or you have a legal right to removal, write to support@worldwires.org with the link. Items are removed from this index on request; the original publisher is not under our control.

Retention

Items stay in the index for 30 days after publication, article text for 14 days, then they are deleted. Nothing is archived beyond that window.

Privacy

Reading needs no account, no cookies, no third-party scripts, no analytics. Server logs (IP address, requested path) are kept for 14 days for abuse prevention; rate limits are enforced per IP address. Pins, read state and saved views are stored in your own browser only and never sent to us. Legal basis: legitimate interest (Art. 6(1)(f) GDPR) in running and protecting the service.

Accounts (optional)

An account only adds personalisation: chosen sources, themes, followed terms, a daily digest. You sign in with a one-time link sent to your email; there is no password. We store your email address, time zone, digest hour, your themes and terms, the matches they produced (for 30 days) and a log of digests sent (90 days). We do not store IP addresses or device details with the account. One first-party cookie (ttw_session, 30 days) keeps you signed in. Mail is sent from our own mailbox; no mail provider sees your address. You can download everything we hold about you and delete the account, with all its data, from the account page at any time. Legal basis: performance of a contract (Art. 6(1)(b) GDPR) for the account, legitimate interest for abuse prevention.

Impressum

Operated by Brightnext Goods and Services Ltd., London SE1 8BZ, United Kingdom. Contact: support@worldwires.org.