Editorial Standards & Ethics
This is the set of rules our editors hold each other to, and the document we point at internally when somebody wants to publish something we should not. Last updated: August 2026.
The format raises the stakes. A news story that gets a detail wrong misinforms somebody. An answer that gets a detail wrong sends them to a dealership, an airport or a pharmacy on our say-so. Every rule below exists because of that difference.
What a claim has to clear before we publish it
Nothing runs on the strength of another outlet’s report. If a rival has the story, that tells us to go and look. It is not evidence. We go to the thing itself: the recall notice, the filing, the patch note, the study rather than the press release about it.
Every number is checked against its source by the writer and again by an editor. Model years, build dates, prices and the day a rule takes effect bite us most, because they are the parts a reader copies down and acts on. Where two official bodies publish different figures for the same thing, we give both and say which produced which. We do not average them.
On a breaking story the pressure is not to be right, it is to be first with something. The one rule holds anyway. We find who holds the answer and ask. If it is not available yet, the story says so in the first two sentences and names who will confirm it and when.
Sourcing and attribution
Statements are attributed to the person or body that made them, on the record, with the date. Quotes are not tidied beyond removing a stumble. Where a document is the source it is linked, or, if it is not public, described precisely enough that somebody else could request it.
Anonymity is granted rarely and never for convenience. A source can be unnamed when naming them would cost them their job, their safety or their case, not because they would prefer to snipe from behind a curtain. The managing editor approves every unnamed source and knows who they are, and the story says what puts that person in a position to know. A source close to the star is not sourcing. We do not print it.
Independence, gifts and outside interests
We take nothing from the people we cover. No gifts, no hospitality, no hardware to keep, no junket flights and hotels, no early access on the understanding that the write-up will be warm. Review units go back or are bought at retail, so we sometimes cover a launch from the stream rather than the room.
Staff and regular contributors declare shareholdings, outside paid work and family interests that touch their beat, and hand the coverage over where the overlap is real. Nobody covers a company whose stock they hold. Staff vote and are otherwise private citizens, and do not campaign for, donate to or publicly endorse a party or candidate while they work on political coverage. The owners are held to the same rules, set out on our ownership and funding page.
News, analysis, opinion and paid content
Most of what we publish is reported news that answers a question. Analysis is labelled analysis, and may reach a conclusion as long as the evidence is in the piece. Opinion is labelled opinion and carries an author. There is no unsigned editorial line here, because there are not enough of us for anybody to hide behind an institutional voice.
Paid content is labelled at the top of the page, in the byline position and in the title, and it names who paid. It is not written by anyone who reports for us, stays out of the question index and the homepage lead, and may not borrow the shape of an answer. The commercial rules in full are on the advertise with us page.
Where the money stops
Advertisers get no preview, no veto and no warning that we are working on something about them. Nobody who sells advertising commissions, edits or approves a story. When an advertiser appears in coverage, the relationship is disclosed in the piece.
A campaign has been cancelled over a story we published. The story stayed up, and we plan the finances assuming it happens again. The formats that perform best are usually the ones we turn down: the affiliate-stuffed answer, the ad unit dressed as a search box.
Telling people before we publish about them
If a story criticises a person or an organisation, we put the substance of it to them beforehand, not a vague request for comment. That means the specific claim, the specific document, and a deadline realistic for the answer we want. A press office that needs to speak to a lawyer gets longer than one confirming a date.
We publish whether or not they reply, and we say what we asked and when. A reply that arrives afterwards is added to the story with the time it landed. No comment is reported as no comment, and silence is reported as silence, because those are different things.
Plagiarism and fabrication
Both end the working relationship. Passing off somebody else’s reporting as ours, inventing a quote, inventing a source, or inventing a detail to close a gap are the only things on this page with an automatic outcome. Where published work is affected we name the conduct on the story rather than calling the piece amended.
Following another outlet’s scoop is allowed and credited at the first mention, with a link in the sentence rather than parked at the bottom. If somebody else got there first, saying so costs us nothing we would want to keep.
How we use AI tools, and where we do not
We use them for a few narrow jobs. Machine transcription of interviews, which an editor checks against the audio. First-pass translation of a document, verified by somebody who reads the language before anything is quoted. Clustering the query log by which readers are asking the same thing. Spelling and grammar checks.
We do not publish machine-generated articles, machine-written quotes or machine-invented sources, and we do not let a model tell us what a document says instead of reading the document. No image here is generated by a model without a label saying so. Every published word is the responsibility of the named writer and the editor who cleared it, and the tool said so is not a defence, any more than the press release said so is. If we use these tools for anything beyond that list, this section changes first.
Who we ask
Our questions come out of a query log, so they come from the people affected rather than the people with a communications budget. The sourcing has to follow. If a story about a fare rule quotes only the airline and an analyst, we have not answered it. The person holding the ticket is the one who typed the question.
So we go after the affected owner, the patient, the passenger, the player and the small supplier, not only the people who return calls quickly. We keep a rough count of who ends up quoted, and when a month reads as press offices and the same four analysts, somebody says so at the weekly meeting. The count is a check on the laziness that convenience produces.
Bylines, and where a complaint goes
Every story carries a named author who is a real person, and the editor who cleared it is on file even where they are not printed. There are no house pseudonyms and no invented bylines. If you need to reach a writer, the desk passes it on.
Everything arrives at support@faq-ans.com, which the editors read and sort themselves. If you think a claim is wrong, our corrections policy sets out what changes on the page and how quickly. How a claim was tested in the first place is in our fact-checking policy, and we work to the IFCN principles. A complaint about an editor’s decision goes to the managing editor, who answers in writing.
