The AI writes it. A human doubts it. Here is what that means in practice.
What the AI does
We use Anthropic’s Claude, an AI language model, to draft our briefs. It reads a published news report and rewrites it as a short, structured summary in plain language, with separate boxes such as What Happened, Who is Affected, Why it Happened and Key Updates.
What a human does
- A person reviews every draft before it is published.
- A person decides whether it goes live, needs changes, or is dropped.
- A person handles corrections and reader reports.
What we do not do
- We do not publish AI text without human approval.
- We do not invent quotes, sources or details.
- We do not present AI-written briefs as original reporting. Each one links to its source.
Limits
AI can misread a source or leave out context. That is why every brief links to the original report, and why we want to hear about errors. Report one on our corrections page.