How to Fact-Check a Viral News Story Before You Share It

A police incident, a burning building, a politician caught mid-sentence. Someone you vaguely know has shared it with the words "please share widely", and your thumb is already hovering over the button. It takes four seconds to pass it on and about four minutes to work out whether any of it is true. Here is how to spend those minutes well.

Find the original post, not the version in front of you

Screenshots are where context goes to die. They strip out the username, the timestamp, the replies and the community note, leaving a confident caption attached to an unverifiable image. The first job is to get back to the original post.

• Search the exact caption in quotation marks. Search engines and the platform's own search will usually turn up the earlier version if one exists.
• Look for the earliest instance, not the most popular one. The post with 40,000 shares is often the fourth or fifth retelling.
• Check who posted it first. Does the account belong to the organisation it claims to represent? Is it a fan account, a parody account, or a handle that has renamed itself to look official? Many fake breaking-news posts come from accounts whose bios quietly say "not affiliated".
• If the trail dead-ends at an account whose entire output is screenshots of other people's posts, you have found a relay station, not a source.

Interrogate the image properly

Reverse image searching is the single most useful habit here, and it takes seconds. Right-click or long-press the picture and run it through Google Lens, TinEye or Bing Visual Search. If it has appeared online before, you will often find it in a completely different context: a different country, a different year, a different disaster.

Crop before you search. A full frame can be too generic; a tight crop around a shop sign, a numberplate, a distinctive uniform or a road marking will narrow things down fast. Then look closely at the picture itself. Warped lettering, hands with too many fingers, shadows falling in two directions and reflections that do not match the scene are all signs that an image has been generated or heavily edited. Grainy pictures are not automatically old, and clean ones are not automatically new.

Video is harder. Take a few frames as stills and reverse-search those, and search for any distinctive audio — a siren, a chant, a newsreader's voice. On YouTube, check the upload date rather than the date printed in the thumbnail, and look at the description for a credit.

Check the date, because that is where most fakes live

Recycled footage is the most common form of viral misinformation, and it works precisely because the footage is real. A genuine clip of flooding, rioting or a hospital corridor resurfaces years later with a new caption. Before you accept a date, ask:

1. When was this uploaded, according to the platform's own timestamp — not the caption?
2. Does anything in the frame contradict the claimed date? Snow in August, leaves on the trees in February, a shop that closed last year, a car model that did not exist yet.
3. Does the weather on the day match what the clip shows? Weather records are free and public.
4. Is the caption phrased in a way that dodges the question entirely — "happening now", "just in", "breaking" with no place and no time?

If a post cannot tell you where and when, that is not a small omission. It is the whole problem.

Read past the caption and the headline

Captions routinely claim more than the article does, and headlines routinely claim more than the story beneath them. Open the link and read down to the point where the claim is actually supported. Often you will find that the source is "social media users say", that a quote has been trimmed mid-sentence, or that a figure is a percentage of something tiny.

Watch the wording too. "Police are investigating", "a man has been arrested" and "a man has been convicted" describe three very different situations, and the gap between an allegation and a court finding is where a lot of viral claims quietly fall apart. If a claim touches on health, money or the law, treat a social post as a prompt to check an official source or a qualified professional — not as advice in itself.

See who else is reporting it — independently

Three outlets carrying the same story is not three sources if two of them are quoting the third. Look for reporting that adds something: a named local reporter on the ground, a statement from an official body, eyewitnesses with detail that could only come from being there. Check whether the story has since been corrected or walked back; newsrooms publish corrections quietly and late.

Be sceptical of screenshots styled to look like a broadcaster's breaking-news alert. Those graphics are easy to copy. Go to the broadcaster's own website or app and see whether the alert is there.

Before you press share

Run the claim through this short checklist. If it fails any line, hold off.

• Can I find the original post, image or document?
• Do I know when it happened, and does the date hold up?
• Does at least one independent source — not a copy of the same story — back it?
• Am I sharing it because it is true, or because it confirms something I already believed?

Then think about what sharing actually does. Adding "unconfirmed" or "if true" does not neutralise a false claim; it usually spreads it further, because people remember the claim and forget the caveat. If you want to push back, reply to the person who posted it with what you found, or share a correction instead of the original. And if you have already passed something on that turned out to be wrong, delete it and say so plainly — a short correction travels further than most people expect, and it costs you nothing.

Photo: Charlotte May / Pexels