Key points
- Misinformation refers to something untrue circulating as if it were true. Deliberate fabrication and an honest mistake passed along both end up mixed together in it.
- Generation tools becoming common has made the cost of producing a convincing bit of text or a photo drop substantially, while the cost of checking it stayed where it was. How large a threat that amounts to is something researchers see differently.
- What tends to spread well shares a few traits: surprising, infuriating, short, and inconvenient to verify.
- A correction usually doesn't travel as far as the original did. Reaching someone who already saw the first version is much harder than reaching them the first time.
- The most reliable thing an individual can do is track down the original, check the date and source, and hold off on passing it along until it's checked.
The analogy
A transit display reading "arriving in 3 minutes" makes everyone waiting settle in and get ready. If the screen is wrong, there's no way to tell just by looking. The number sits there crisp, the screen looks perfectly normal.
Someone says "it's coming soon," and the line stirs. Someone who was about to head to a different stop turns back, and that person passes the same line to somebody else. One display ends up reshaping several people's evenings.
Even once the screen gets corrected, much later, things don't snap back. Whoever already left doesn't see the corrected screen, and the message that got passed along keeps rolling regardless of the correction. Misinformation making its way around looks a lot like that display.
In detail
The more convincing it feels, the further it travels
Look at what actually spreads and it's how strongly something hits an emotion, more than whether it's true, that drives how far it goes. A story that's shocking, infuriating, or unjust gets a finger moving before anyone checks it. A flat, accurate explanation just doesn't travel nearly as far.
Format matters too. A short one-liner, a bold flat claim, a single number, one photo, anything shaped to be passed along whole tends to circulate well. A sentence loaded with qualifiers and caveats tends to get its edges trimmed off as it's passed along, and the trimming makes it sound more certain than it was.
Verification gets especially loose around anything that already matches what someone believed going in. The instinct to check tends to switch on only for an unfamiliar claim, and barely switches on at all for a familiar one.
Making it got faster; checking it didn't
It used to take real effort to produce a convincing fake. Touching up a photo took time, and writing a long, convincing piece of fiction took a person sitting down to write it. Now that effort has dropped substantially. Taking the same claim and spinning up hundreds of versions with only the wording changed isn't hard either.
The cost of checking, on the other hand, stayed exactly where it was. Tracking down the original, lining up the date, and comparing it against other sources still takes a person's time. Many people see the gap that's opened up between how fast something can be made and how fast it can be checked as the biggest change. Some researchers argue the effect is smaller than it looks, since even when many fakes are made, only a limited number of people go looking for them and believe them, so views differ on how large the threat is.
When the same story shows up worded slightly differently in several places, people tend to read that as several independent people each having confirmed it. In reality it can just be one thing copied out several times.
A correction doesn't travel as far as the original
A correction usually shows up late and travels less. There's nothing new about it, it doesn't hit any emotion, and it comes wrapped in qualifiers that make the sentence longer. Only a fraction of the people who saw the original ever run into the correction at all.
Even how a correction gets written matters. It used to be a common worry that restating a false claim would cement it further, but later research suggests that backfire effect is rare. Leading with the accurate explanation, flagging the false claim clearly as false and mentioning it only once, and adding why the misunderstanding happened in the first place tends to stick better.
A belief that's already settled in tends to stick around even after the evidence under it has fallen apart. Which is why stopping something before it spreads is a far better use of effort than fixing it after.
An order for checking
None of this needs elaborate tools. Just an order to follow. Start by tracking down the original. Skip the version that's been copied and passed along, and find where it first appeared, and the surrounding context that comes with it often changes what it means on its own.
Next is the date and the place. Old material passed off as something recent happens constantly. For a photo, running a reverse image search often shows how long it's actually been floating around online. Finding no record at all, though, doesn't mean the photo is genuine.
Last is comparing it against other sources. If something is genuinely a big deal, several outlets should each be covering it independently. If it only shows up in one place and everywhere else stays quiet, that silence is itself a signal.
Not spreading it counts as a response too
Simply holding off on passing something along until it's checked already does a lot of the work. Even a post that quotes something false in order to criticize it ends up spreading the original claim along with the criticism.
An AI's answer alone doesn't finish the checking, either. A conversational AI builds its answers by continuing plausible-sounding sentences, so it can invent a source that doesn't exist. Even when search is attached and it shows sources, whether the source is actually real, and whether it really says that, is still something a person has to go click and check.
More precisely
Splitting the term makes it more precise. Passing something along without knowing it's false is called misinformation in the narrow sense; spreading it while knowing it's false, in order to deceive or do harm, is called disinformation. This article uses "misinformation" for both together, though the word is sometimes used only for the unintentional kind. What actually circulates is usually a mix of both. The person who first made it may have had an intent, but most of the people who pass it along believe it's true.
The analogy breaks down in one place: a transit display has one operator, so a fix reaches everyone who looks at the same screen. On the internet, everyone's screen is separate, assembled by a different feed for each person, so the same correction doesn't reach everybody who saw the original. And a display's errors are usually honest accidents, while what circulates online has some material mixed in that was made on purpose.
Leaning on detection tools has limits too. A tool that checks whether a photo was AI-generated only hands back a probability, never a verdict, and there's plenty of misinformation in a form no tool can touch at all, like a real, unaltered photo carrying a fabricated caption.
Try it yourself
Common misconceptions
It's easy to think a fake is sloppy enough to spot on sight, but actually a growing body of research shows people's accuracy at telling real from fake sits close to a coin flip.
It's easy to think asking an AI settles whether something is true, but actually it can invent a convincing source that doesn't exist, so its answer alone doesn't finish the checking.
It's easy to think pointing out that something is wrong gets it corrected, but actually a correction usually doesn't travel as far as the original did, and a belief that settled in first can stick around.
One-line summary
Misinformation is one convincing signal moving people to act, and since making one got faster, the habit of checking the original and the date before passing anything along matters more than ever.
Spotted an error or have a better analogy? Suggest an edit · Last updated 2026-10-18