Scam Watch
A short brief each morning on what official sources published about scams in the previous day, with every item linked back to the agency that issued it.
How these are made, and their limits
Each morning a job fetches the newsrooms and alert feeds of regulators, police forces and prosecutors, then a local language model writes a brief from only those items. Two guards run before anything publishes: every URL in the brief must appear in the fetched set, and every scheme tag must correspond to a real page on this site. If either check fails, the brief is retried once and then skipped — the site never publishes a brief whose links it could not verify.
These are marked auto-generated and human review pending. The linked source is always the authority; the summary is a pointer to it. Analysis and anything with a claim in it goes in the blog, which is written by a person.