You know that little rush ? Someone posts a screenshot, the caption is furious, and your thumb is already halfway to the share button. I’ve done it. Twice, publicly, and both times I had to go back and delete the post, which is a special kind of embarrassing when your aunt has already commented “unbelievable !!” underneath. So this article is basically the checklist I wish I’d had. Five steps, no fancy tools required, and honestly most of them take less than three minutes. By the end you’ll know how to verify information before sharing it, which red flags to look for, and where to check when you’re genuinely not sure.
Here’s the boring habit that changed things for me : before I share anything, I open two or three other outlets and see whether anybody else is running the same story. A general news site like https://slnews.fr, a big wire agency, one specialist source – and if literally nobody else has picked it up, that silence tells me something. Real events leave traces. A big claim that exists in exactly one place, with no source, no date, no name ? Franchement, that’s usually not a scoop. That’s a rumour wearing a suit.
Step 1: Stop. Notice what the post did to your body
This one sounds soft and it’s the most important. Disinformation isn’t designed to inform you, it’s designed to move you. Rage, fear, smugness, that warm “I knew it” feeling. Researchers at MIT looked at years of Twitter data back in 2018 and found false stories travelled faster and further than true ones, partly because they were more novel and provoked stronger reactions like surprise and disgust. Which, when you think about it, is pretty depressing.
So the rule is simple : the stronger your reaction, the slower you go. If your jaw is tight, you’re the target audience. I actually count to ten now. Not because I’m zen – I’m not – but because those ten seconds are the difference between me being a reader and me being a distribution channel.
Step 2: Find the original source, not the messenger
Ninety percent of the stuff in your feed is a copy of a copy of a copy. Someone screenshots an article, someone else screenshots the screenshot, and by the third round the headline has quietly changed a word or two. That word is often the whole story.
So ask : who said this first ? Not “who sent it to me”. Who originally published it. Look for a named author, an actual outlet, a link that goes somewhere real. If it’s a study, does anyone name the journal ? If it’s a quote from a politician, is there video, a transcript, an official statement ? A claim with no traceable origin is not information. It’s folklore.
Little trick I use : copy a distinctive phrase from the post – five or six words, in quotation marks – and search that. If the claim is real, you’ll usually land straight on the original. If it’s fake, you’ll land on a pile of identical posts and nothing else.
Step 3: Check the date. Seriously, check the date
Perso, I think this is the most underrated step of the five. A huge share of what people call “fake news” isn’t fake at all – it’s true, but it happened in 2017. Old floods, old fires, old protests, old resignations, recycled during a new crisis because they get engagement.
Look at the publication date. Look at whether the article has been updated. If it’s a photo, ask when it was taken, not when it was posted. Ever noticed how the weather in a “breaking” photo doesn’t match the season ? Leafless trees in an August story. Winter coats in June. Once you start looking for that stuff you can’t unsee it.
Context matters just as much as timing. A real quote cut off two sentences early can mean the exact opposite of what the speaker said. That’s not a mistake, usually. That’s the technique.
Step 4: Interrogate the image or video
Photos feel like proof. They aren’t, and they haven’t been for a long time. The good news is that reverse image search is free and takes about twenty seconds.
- Google Lens or Google reverse image search – drop the image in, see where else it’s appeared and when.
- TinEye – good at finding the oldest version of an image online, which is exactly what you want.
- Yandex Images – often surprisingly strong on faces and places others miss.
- The InVID-WeVerify browser plugin – made for journalists, splits videos into keyframes so you can search them frame by frame. Free, and it feels like a cheat code the first time you use it.
With AI-generated images the old advice about counting fingers is basically obsolete now, so I’d focus elsewhere : text in the background that dissolves into nonsense, reflections that don’t line up, jewellery or teeth that go slightly wrong, and above all – does any credible outlet have a different photo of the same event ? A real event gets photographed from several angles by several people. A generated one exists exactly once.
Step 5: Cross-check sideways, not downwards
Most of us “verify” by reading the suspicious article more carefully. That’s backwards. Professional fact-checkers do the opposite – they leave the page immediately and go find out who’s behind it. It’s the core idea of the SIFT method developed by Mike Caulfield : Stop, Investigate the source, Find better coverage, Trace claims back to the original.
Practically, that means opening a few tabs : two or three unrelated outlets, ideally with different editorial leanings, plus the fact-checking services that do this full time – AFP Factuel, Reuters Fact Check, Snopes, Full Fact, PolitiFact, depending on the topic and country. And if you want to see whether a page has quietly changed its wording since publication, the Wayback Machine at archive.org keeps the receipts.
Independent confirmation means genuinely independent. Ten sites all citing the same anonymous Telegram post is one source, not ten.
The red flags I check almost automatically now
No author name. No date. A URL that’s almost a famous outlet but with an extra letter or a weird domain ending. ALL CAPS in the headline. “They don’t want you to see this.” A statistic with no institution attached. An emotional call to action – “share before it’s deleted !” – which is the oldest trick in the book, because urgency kills verification. And screenshots of text with no link back. Screenshots are the natural habitat of the unverifiable claim.
What if you still can’t tell ?
Then don’t share it. That’s it, that’s the whole answer. There is no obligation to have an opinion on everything within four minutes of hearing about it. The Reuters Institute’s annual Digital News Report has been showing for years now that a large share of people – close to six in ten in recent editions – say they worry about telling real from fake news online. So you’re not alone in the doubt, and doubt is honestly the healthiest state to be in.
And if you got it wrong ? Correct it out loud. Delete the post, say why, don’t bury it. I find people respect that far more than they mock it. Nobody expects you to be a newsroom. They just expect you to care whether it’s true.
Five steps : pause, find the source, check the date, test the image, cross-check sideways. Three minutes, most of the time. Which, given how much damage a bad share can do to someone’s reputation or to a family group chat, feels like a pretty reasonable price to me.