Editorial Standards & Ethics

These are the rules our newsroom works to. They describe what happens to a story between the moment a reporter files it and the moment it appears here, and what happens when one of them is broken.

Last updated: August 2026.

Accuracy before speed

Nothing goes out because it is probably true. Every factual claim has to be traceable to something we have seen: a document, a dataset, a person on the record, a statement we can attribute. If a reporter cannot say where a number came from, the number comes out.

The hardest test is a developing story. A quake off Sumatra, a ferry overdue, a minister resigning by mid-morning, and the first hour brings figures wrong in several directions. We publish what we can attribute and mark the rest unknown, including the casualty count, which we leave open rather than lift from a screenshot. Tolls get revised, so we name the agency and the time.

We read the original Indonesian, not an English summary. Aggregation garbles Indonesian officialdom predictably, and one mistranslated term can move an evacuation zone fifty kilometres. We know because we did it. Early on we published an evacuation radius attached to the wrong regency, lifted from an English write-up of a briefing nobody here had read.

Sourcing, attribution and anonymity

Named sources on the record are the default and carry most of our reporting. Where we quote a document, we link it or describe it precisely enough that you could request it yourself. Where we take something from another outlet we credit them by name with a link, rather than laundering a competitor’s scoop into “reports said”.

Anonymity is granted for a reason, never for convenience. The test is whether the person faces a real cost, losing a job in a ministry or a contract, and whether what they say is corroborated elsewhere. We do not print “sources close to” a company or a celebrity as cover for a publicist briefing us about a client. Every anonymous source is known to an editor, who approves the arrangement, and the story says why the name is withheld.

Independence, gifts and political activity

We do not take press trips. No free flights, no hotel nights, no tourism board junkets, no review unit kept afterwards. That is how travel and lifestyle coverage is normally paid for here, and refusing it costs us money and access. Anything unsolicited goes back.

Nobody on the desk owns shares in a company they cover, and holdings in sectors we cover are declared to our managing editor. Editors do not consult for companies, ministries or campaigns, or take speaking fees from bodies on their beat. Family in government, or inside a company on a reporter’s beat, is disclosed and the beat moves. Staff vote and hold private opinions. They do not donate to parties, hold membership, or run partisan accounts while covering politics, and the paid influencer placements offered every election season are refused. Our ownership and funding page says the same at company level.

Reporting, analysis, opinion and paid work

Reporting carries no view of our own. Analysis is written by someone who covers the subject, is labelled analysis, and may reach a conclusion from evidence laid out in the piece. Opinion is labelled opinion and argues a position the writer holds. If you cannot tell which you are reading, our labelling has failed.

Commercial work is labelled at the top of the page and in the feed, names the sponsor, and never runs under a reporter’s byline. A reporter who covers a beat as news does not write sponsored copy on it. Ministry cooperation-agreement packages, where a government body buys a run of friendly articles, are refused rather than labelled. Our advertising terms list what we turn away.

The wall between the desk and the money

No advertiser sees a story before publication. None is warned one is coming, none gets a preview or a veto, none gets a sentence altered. The commercial side has no access to the editing queue, and nobody who edits is measured on advertising outcomes.

The desk is told which companies advertise with us for one purpose, so the relationship is disclosed inside any story about that company. An advertiser’s complaint runs through the same process as a reader’s. We do not accept paid takedowns. Those requests arrive most weeks with a figure attached, and the figure has never changed the answer.

Telling people we are writing about them

Anyone facing criticism is contacted before publication, told specifically what we intend to report, and given a real chance to answer. Not a message sent twenty minutes before we publish. With a company or an agency we go to the office that can respond, say when we need the reply, then publish the answer or the fact that we asked and got nothing.

A response that arrives after publication is added to the story, with a note saying we added it. Silence is reported as silence, never as agreement. On a fast story where waiting would leave readers exposed to something urgent, we publish, keep asking, and say in the piece that we have no answer yet.

Plagiarism and fabrication

Both end the relationship. Copying another outlet’s sentences, lifting a description from a press release without attribution, passing off a translation of someone else’s reporting as our own, or recycling our published copy as new reporting are not slips to be managed. Inventing a quote, a source, a document or a scene is worse.

Where it happens we take the work down or rewrite it properly, publish a note saying what was wrong, credit whoever we took from, and part company with the person responsible. If a freelancer did it, everything else we bought from them is re-read.

Where we use AI, and where we do not

Plainly, then. We use these tools for mechanical work and not for reporting. They transcribe Indonesian audio from a press conference, give us a first pass at a long ministry document, tidy our clumsy sentences, and find the paragraph we need inside a four hundred page PDF. A person checks every output against the original, because a transcription tool mishears place names and a translation tool smooths an evasive answer into a clear one.

We do not publish model-generated text as reporting, do not let a tool summarise a story nobody here has read, and do not use synthetic images or audio in news coverage. Nothing appears under a byline unless a human reporter gathered it and a human editor read it line by line. If a tool ever does more than the work described above, the story says so at the top. A mistranslation that reaches print is our error, not the software’s.

Whose voices end up in the story

A story about a fuel subsidy change that quotes only economists in Jakarta is an incomplete story. We look for the people a policy lands on, which usually means going outside the capital, and we work at reaching women in sectors where every official spokesperson is a man. That is slower, and it is the first thing to suffer when we are rushed.

We interview in Indonesian and in regional languages and translate carefully, instead of quoting only the English-speaking officials who are easiest to book. Where a story concerns a region, we want someone who has been there. Where our knowledge runs out, the piece says so.

Bylines, and holding us to this

Every story carries the name of the person who wrote it. Where several reporters contributed, all are named, and where a piece really is the work of the desk it runs under our newsroom byline rather than an invented person. A withheld byline happens only where naming the writer would put them at risk.

You can reach whoever wrote a story at support@wisatahits.blog, an inbox our editors read and sort themselves. If you think a story broke one of the rules above, name the rule and quote the sentence. Our corrections policy says what happens once an error is confirmed, and our fact-checking policy describes the checks a story goes through first.