Misleading posts spread because they are built to. They are surprising, emotional and easy to pass on, and by the time a correction appears the original has already been shared thousands of times. You cannot verify everything you see. But a short routine, taking a minute or two, catches a large share of the false material that circulates.
Stop before you share
Stop. The pull toward sharing is strongest in the first few seconds, when a post confirms what you already believe or makes you angry. Strong feelings are a good signal to slow down, because content designed to provoke is often content designed to spread.
Researchers who study online reasoning have turned this into short checklists. One popular version, called SIFT and developed by the digital literacy researcher Mike Caulfield, runs: stop; investigate the source; find better coverage; and trace claims, quotes and media to the original context. The first letters of those four steps spell SIFT. The sections below follow that sequence.
Who is behind it?
Check who posted the claim before you evaluate the claim itself. A name or a handle is not enough. Look for what the account or site has published before, who runs it, and whether anyone independent describes it.
The most reliable technique has a name: lateral reading. Instead of studying the site’s own “About” page, open a new tab and search for what other sources say about the site. Professional fact-checkers do this routinely, and studies by the Stanford History Education Group, now the Digital Inquiry Group, found that fact-checkers reached accurate conclusions about unfamiliar websites faster than people who stayed on the page.
Also check for impersonation. Look-alike accounts copy logos, add a verified-looking symbol and change a single letter in the name. A claim attributed to an agency or a newspaper should be findable on that organization’s own website.
Does anyone credible confirm it?
Search for the central claim in plain words and see who else reports it. A major event leaves a trail: wire services, local outlets, official statements and court or agency records. If only anonymous accounts and a handful of unfamiliar sites repeat it, treat that as a warning.
Fact-checking organizations that follow a published code of principles can help. The International Fact-Checking Network, run by the Poynter Institute, maintains a code that signatories commit to, including transparency about sources and corrections. Searching a claim alongside the word “fact check” often turns up an existing review.
Health claims deserve the same care. A post about a vaccine, for example, usually mixes together steps that different agencies handle, which is why it helps to read how vaccines are tested and approved before judging what a headline implies.
A claim that is partly true is common. A real photo with a false caption, a real quote with the surrounding sentences removed, or a real statistic compared to the wrong baseline can all pass a casual check. This is why the next step matters.
Tracing it to its source
Many viral items are old material in a new setting. To trace one:
- Reverse image search. Upload the image or paste its address into a tool such as Google Lens or TinEye. If the same picture appears years earlier, attached to a different event, the caption is wrong.
- Check the date. Sharing an old article as though it were current news is one of the most common errors. Look for a publication date and be wary of screenshots with none.
- Find the full quote or video. Clips can be trimmed so a sentence means the opposite of what was said. Look for the longer recording or the full text of the speech.
- Find the data. A chart without a source is a drawing. Look for the agency, study or survey behind a number, and check its date and what it actually measured.
AI-generated images and text add another layer. Generated pictures can look convincing, and a text model can produce a confident but false paragraph, so a smooth, plausible surface is no longer evidence that something is real. It helps to know what large language models can and cannot do when you see text that seems to have no clear author.
Warning signs worth heeding
- A demand to share immediately or “before it gets deleted.”
- A claim of a cover-up that nobody else seems to report.
- Heavy emotion and few verifiable details: no names, no dates, no place.
- A screenshot of a headline rather than a link to the article.
- Numbers that are strikingly round, or that conflict with official figures you can look up.
None of these proves a post is false, but all of them justify a closer look.
If you already shared it
It happens to everyone. Delete the post or add a correction, and say plainly what was wrong. A short note works better than silence, because people who saw the original may never see a later fix. Replying to the people who reshared it is worth the effort when it is easy to do.
The same discipline applies to polls and statistics, which are often quoted without context; how opinion polls work and what the margin of error means is a useful companion for anyone who shares numbers. And it connects to a bigger skill: knowing how to tell news reporting from opinion, so you know what kind of claim you are being asked to believe.
Sources: Stanford History Education Group; Digital Inquiry Group; Poynter Institute; International Fact-Checking Network
