Behind the Famous publishes reference pages about the families of public figures. Most of the people we write about never sought attention. They are somebody’s husband, somebody’s daughter, somebody’s father. The only reason their name is searched at all is the person they are related to.
That shapes everything about how these pages are made. This page explains the rules, so you can judge whether to trust what you read here.
Where the facts come from
Every specific on a page — a date, a place, a job, a relationship — comes from a structured record we assemble before any of the page is written. That record draws on two things: typed relationship data from Wikidata, and the sentences in Wikipedia articles that name the person, together with the references those sentences cite. Where a Wikipedia sentence rests on a newspaper or magazine report, it is that report we list in the Sources block, not the Wikipedia article.
We do this because the open web on these people is mostly machine-written and contradicts itself. A search for the spouse of a well-known politician will return a dozen pages that disagree about whether the couple are still married, several of which name the wrong person entirely. Starting from a structured record is the only way to avoid inheriting that.
What we will not publish
- Anything about a child. If the subject of a page would be under eighteen, the page is not made. There is no volume of search traffic that changes this.
- Crime, illness, addiction, fertility, sexuality, affairs or cause of death — unless the public record establishes it and the person it concerns, or their family, put it on that record themselves. Where we do report such a thing, we attribute it and we do not embellish it.
- Estimated net worth, salary or income. These figures are invented almost everywhere they appear. We publish none.
- Home addresses, schools, employers of private individuals, or personal social media accounts.
- Hedged claims. No “reportedly”, no “sources say”, no “it is believed”. If we cannot stand behind a sentence, it is not on the page.
How the pages are written
The prose on these pages is drafted by a language model that is given the structured record and nothing else, then checked automatically against that record before anything is published. The check removes any sentence containing a date, a figure, an age or a claim the record does not support, and any sentence that denies something the record establishes. Pages that lose too much to that check are not published at all.
We say this plainly because we think you are entitled to know. A model that is allowed to supply facts about a private person will eventually supply a wrong one, and a wrong fact about a named individual is not a typo. The point of the structured record is that the model never gets to supply a fact in the first place.
What we say when we do not know
Most of these people have thin public records. Where a detail is missing, the page says so rather than estimating. Every page carries a section setting out what has not been reported — usually a birth date, a career history, how a couple met. We think that section is the most useful thing on the page, because it is the part no other site will tell you.
Corrections
If a page about you or someone you know is wrong, we will fix it. If you would rather the page did not exist, say so and we will take it down. See corrections and removal requests.