For publishers

How Raven Wire uses your reporting, and how to change that. We act on any request within 48 hours.

What we take

From your public RSS or Atom feed we take the headline, the link, the publication time and your outlet’s name. We keep a feed’s short description only for public-domain sources and publishers who have given permission, and never for wire copy. For everyone else we read it once, note two facts (whether it is wire copy, and the places it names), and discard it.

What we show

  • Your headline as the text of a link to your article, with your outlet’s name and the time. Every click goes to your site.
  • Our own verdict and counts: how many independent newsrooms report the story, in how many countries.
  • Never your article text, ledes, descriptions, images or thumbnails: not on our pages, not in alerts and not in our public API. We don’t get around paywalls.

AI

Our AI writes short summaries across several outlets’ reports about the same event. They are labelled “AI-generated” wherever they appear, including in our data. How we use AI. What the AI reads from each publisher:

  • Public-domain sources and publishers who gave permission: the headline and up to 400 characters of the feed’s description.
  • Everyone else: headlines only.
  • Publishers whose terms bar AI use, that declare a TDMRep reservation, or that opt out of AI: nothing, not even the headline. Their reporting still counts toward how many newsrooms report a story.

We never train or fine-tune AI models on your content.

How to block us or opt out

Our fetcher identifies itself as RavenWire/0.1 (+https://theravenwire.com). A few feeds that reject unknown clients see Mozilla/5.0 (compatible; RavenWire/0.1; +https://theravenwire.com). It always contains RavenWire.

  • Stop us fetching your feed: add this to your robots.txt. We check it every 6 hours, for RavenWire (and our earlier name, AlertNewsTicker) or else *, and stop fetching any feed it disallows.
    User-agent: RavenWire
    Disallow: /
  • Keep your content out of our AI only: declare a TDMRep reservation (tdm-reservation: 1 in /.well-known/tdmrep.json or as an HTTP header on your feed). We check it every 6 hours. We can still list and count your headlines.
  • Anything else (removal, AI opt-out, a different name or attribution, a correction): email publishers@theravenwire.com. We act within 48 hours and confirm by email. Once we record your request it applies on our next run, within 10 minutes.

When we stop fetching a feed, headlines we already list stay until their stories age out: after 7 days for most stories. Stories that reached alert level L3 or higher are kept longer as a record of major events.

Status, November 17, 2026

  • In place: everything on this page. Headline, outlet and link only; no publisher text or images on our pages, in alerts or in our API; descriptions kept only by permission; AI input by permission; robots.txt and TDMRep checked every 6 hours; opt-outs applied on the next run; AI summaries labelled; no AI training.
  • Descriptions we had stored earlier from other publishers have been deleted.
  • Publishers whose feed terms bar this kind of use aren’t included until they give permission.

Sources we read

The full list, with each source’s tier, ownership and country, is on the Method page. How we score and label stories: Editorial standards. Other questions: Contact.

Last updated November 17, 2026.