AI policy
We use AI as a tool, never as a journalist. Human editors remain responsible for everything we publish.
What we allow
AI may assist with translation between French and English, copy-editing, summarising public documents, suggesting headlines or structuring research. These are drafting aids, reviewed by a human before anything is published.
What we never do
We do not publish AI-generated articles as if they were reported by a journalist. We do not treat an AI model as a source of facts: figures, quotations and claims are always traced back to verifiable sources. We do not generate fake photographs of real events.
Human responsibility
Every article is validated by a named editor who is accountable for its accuracy, its sources and its tone. AI assistance never removes that responsibility.
Disclosure
When AI assistance has a significant impact on a piece of content, we say so. Every article remains checked, decided, signed and owned by the newsroom.
Access for AI engines
We want our verified content to be quotable by search and AI engines. Our articles are structured for clarity — a short answer up front, key takeaways, cited sources and a verification date — and our crawl rules allow reputable engines to read the content that matters.
