Fact-Checking Policy
What this policy covers
This page describes how INCLUDED NEWS verifies information before publishing it and how we handle claims that resist verification. It applies to every desk on the site: a business story about earnings is checked the same way as a breaking report from a disaster zone, with methods adapted to the material but the standard held constant. Our Editorial Standards page sets the broader rules; this page is the checking machinery itself.
Verification before publication
Every article goes through the same gate before it appears on includednews.com:
- each factual claim is traced to a source: a document, a dataset, a named statement, or a wire report from an established agency;
- names, titles, places, dates and numbers are checked character by character, because these are where most published errors actually live;
- quotes are checked against the recording, transcript or original text they came from;
- a second set of eyes reviews stories making significant or damaging claims before publication;
- anything that fails these checks is cut, attributed and flagged, or held until it passes.
Breaking news and wire discipline
Breaking stories are where fact checking is hardest and matters most, so our rules there are the strictest form of the general policy. We treat early casualty figures, suspect identifications and eyewitness claims as provisional by default, and we attribute them to their sources rather than asserting them in our own voice. We follow wire discipline: a major claim needs confirmation from an established wire agency, an official source, or our own direct verification before we carry it, and a claim carried by many outlets that all trace back to one unverified origin still counts as one unverified source. When the confirmed facts are thin, we publish a short accurate story and expand it, rather than padding with speculation. Developing stories are labeled and timestamped so you can see what we knew when.
What counts as a primary source
Our ranking is explicit. Strongest: original documents (court filings, legislation, regulatory disclosures, published studies), raw data from official statistics bodies, direct statements by participants, and material we have verified ourselves. Next: reporting by established wire services and reputable outlets, credited and linked. Weakest, and never sufficient alone: social media posts, screenshots, forwarded messages, anonymous claims and content from accounts with no verifiable identity. A press release is a primary source only for what an organisation says about itself, never for whether it is true.
When official numbers disagree
Governments revise figures, agencies contradict ministries, and two official sources can publish two different death tolls on the same evening. When that happens we report the disagreement itself: both figures, both sources, and the likeliest reason they differ, rather than picking the more dramatic number. Precision we do not have is precision we do not print.
Statistics, studies and surveys
Numbers get special handling because they carry false authority so easily. We check what a statistic actually measures before repeating what a press release says it measures. We distinguish correlation from causation in our own wording, note sample sizes and funders where they matter, and treat a single unreplicated study as exactly that. Percentage claims are recomputed from the underlying figures when the source provides them, because a surprising percentage is more often a maths slip than a scoop.
Images and video
The most common visual misinformation is not fabrication; it is a real image from a different time or place attached to a new event. Visual material is checked for origin before use: reverse image search, cross checks against known coverage of the claimed event, and scrutiny of details that anchor a scene to a time and place. Synthetic media is an active threat in breaking news, and dramatic unverifiable footage stays out of our coverage no matter how widely it is circulating.
Links are the check on us
Wherever a claim rests on a public source, we link the source itself, not merely a page that mentions it. This is deliberate self exposure: a reader who can click through to the filing or the dataset does not have to take our word for anything. If you follow one of our links and find the source does not support our sentence, that is exactly the kind of failure we most want reported to support@includednews.com.
The same test for every claim
Verification here is symmetric. A claim that flatters a popular position gets the same scrutiny as one that offends it, and a story our readers would love to be true is checked harder, not softer, because wanting something to be true is the oldest hole in any checking process. We hold our own reporting to the standard we apply to officials: if we would demand a source for it from a government spokesperson, we demand one from ourselves.
When a check fails after publication
No gate catches everything. When a published claim turns out to be wrong, the Corrections Policy takes over: the error is fixed promptly, labeled visibly, and logged internally so the checking step that missed it gets strengthened. Readers who catch what we missed are doing this page’s work for us, and the address for that is support@includednews.com with the subject line Correction.
