AI Disclosure
Last updated: June 1, 2026
Draft — pending legal review. Honest first-draft wording, not legal advice.
We openly label AI involvement in our content — for humans and machine-readably. This aligns with the spirit of the EU AI Act (Art. 50, applicable 2026-08-02) and with our trust-first principle.
How we use AI
Language models help read sources and draft summaries, "why it matters" blocks and EN↔UK translations. AI does not decide "what to publish" — a human does.
Human oversight
Every brief passes a human editor before publishing: fact-checks, tone, de-duplication. The editorial team — not the model — is responsible for what is published.
Sources & attribution
Every story links to its primary source. We aggregate and explain, adding value (context, takeaways), rather than passing others’ work off as our own.
Machine-readable labelling
Pages carry structured data about AI involvement; for any generated media we plan C2PA metadata. This lets platforms and answer engines label our content correctly.
Accuracy & corrections
AI can make mistakes. Spotted an error? Email [email] and we’ll fix it and mark the correction. Always verify critical information against the primary source.