Reuse of announcements by search engines and AI assistants

A press release is meant to be shared. Clarsource publishes every announcement in a format that people and machines can read, with its provenance. Here are the rules for reuse.

Clarsource, a service from FinadoryUpdated on 20 October 2026

Why Clarsource is designed for AI too

Search engines and AI assistants have become a common way to find out about a company. To present an announcement accurately, they need three things: the text approved by the issuer, the key data separated from the text, and the provenance of each item of data.

As well as the text, each announcement page therefore contains structured data in schema.org format: issuer and SIREN, date, amounts, locations, people, verification level and version.

What can be reused

Announcements published by their issuers on the public feed are intended to be shared. They may be:

  • indexed by search engines;
  • quoted and summarised by AI assistants;
  • reused by the media and databases;
  • integrated by a developer via the feed or the API.

Conditions for reuse

Reuse must remain faithful to the announcement and its source:

  • credit the issuer and link to the announcement page;
  • do not change figures or assign them a verification level they do not have;
  • use the latest version: a corrected announcement replaces the previous one;
  • take withdrawals into account: a withdrawn announcement must no longer be presented as current.

Example wording: Fictional example “Forges de Vendée, press release published on Clarsource on 12 October 2026”, with a link to the announcement.

What is excluded

  • announcements under embargo or shared via a private link, which are not public;
  • information provided by confidential sources and exchanges with the newsroom;
  • photos, logos and attached documents, which remain subject to their authors’ rights;
  • personal data belonging to press contacts, beyond what is needed to make contact about the relevant announcement.

Technical access

Public pages are accessible to search engine crawlers. For regular or high-volume use, use the access options provided instead:

  • the RSS feed for the feed and each category;
  • the API, which provides the status of each announcement (published, corrected, withdrawn) and its version history.

For use beyond this scope, such as training a model on all announcements, contact us before collecting any data.

For issuers

Publishing on Clarsource means accepting that the announcement may be reused under these conditions. If a reuse misrepresents your announcement, let us know: we do not control third-party services, but we can help you request a correction by providing the authoritative version.

Building a service?

The API provides announcements, their structured data and their status, version by version.

Explore the API

Frequently asked questions

Can an AI assistant quote a Clarsource announcement?

Yes, provided it credits the issuer, links to the announcement and does not change any figures.

Can crawlers see announcements under embargo?

No. An announcement under embargo is not published until the time chosen by the issuer, and an announcement shared via a private link is never public.

How can I tell if an announcement has been corrected or withdrawn?

The announcement page displays its history. The API shows the status and version of each announcement.

Can I use announcements to train a model?

Contact us before collecting data for this purpose. No permission is needed for normal use, such as quoting and summarising an announcement.

Trust and transparency