About NXEWS

What NXEWS is

NXEWS is a news site with no human reporters. An automated newsroom built on Claude reads the public feeds of hundreds of news organisations, agencies and public broadcasters, works out which of them are covering the same event, and writes one story from what they collectively reported.

Nothing here is a copy of somebody else’s article, and nothing here is a scoop of our own. Every story is a synthesis of already-published reporting, written in our own words — and it tells you exactly whose reporting it was built from.

How a story is made

Four steps, repeated around the clock:

1. Collection — feeds from wire services, national broadcasters and regional outlets are pulled continuously and stored as they arrived.
2. Clustering — reports about the same event are grouped across languages and countries, so a story is assembled from several independent accounts rather than a single one.
3. Writing — the newsroom is handed only the facts contained in those reports, and is instructed to refuse to add anything else: no background it was not given, no figure it was not told, no quote it did not receive.
4. Publishing — the story goes out with its sources attached, in every language we publish.

A story that one outlet is carrying is treated as thinner than one that four outlets are carrying, and the front page ranks it accordingly.

Sources and attribution

Every published story carries the list of outlets it was written from, and each of them links straight back to the original article on the publisher’s own site. We would rather send you to them than keep you here.

Where a picture came from is part of that attribution: a photograph on NXEWS is the one the source feed supplied, shown with its credit and a link to where it was published.

Two publication languages

The interface language is decided on the server before the page is built, and right-to-left remains a first-class layout. Article copy is generated only in English and Hebrew.

Every story is written in English and translated to Hebrew at publish time. No other language is offered and no other paid translation can be triggered.

What this can get wrong

An AI newsroom inherits every error in its sources and can add errors of its own. It can confuse two similar events, it can carry a number forward after the original outlet has corrected it, and a machine translation can shift the emphasis of a sentence even when every word in it is right.

That is why there is a sources block at the foot of every story. If something matters to you, follow the links and read the original reporting.

Pictures

No image on this site is AI-generated. Either a source supplied the photograph — and then it appears with a credit and a link — or the site draws an abstract cover from the story’s section and headline, which is meant to look like design and never like a photograph.

Corrections

When a source corrects itself, the story here is rewritten from the corrected material and the update time on the page changes. If you find something wrong, the fastest route to a fix is the original outlet listed under the story.