Published on 2026-07-24
Why false news travels faster than the truth (and no, it's not the bots)
In 2018, MIT researchers traced 126,000 rumour cascades on Twitter and found falsehoods spreading about six times faster than facts. The reason says something about all of us.
Here is a race nobody signed up for. A true story and a false one leave the starting line at the same moment, on the same platform, chasing the same audience. The false one wins. Not by a nose: by laps. That, in one line, is what three MIT researchers found when they ran the largest study of its kind, and their explanation is less comfortable than the one most of us were hoping for.
In 2018, Soroush Vosoughi, Deb Roy and Sinan Aral published a study in the journal Science that traced around 126,000 chains of shared news stories on Twitter, posted between 2006 and 2017 by about 3 million people, 4.5 million tweets in all. Every story in the pile had been checked by six independent fact-checking organisations, so the researchers knew which ones were true and which ones were false. Then they watched how each kind moved.
Falsehood wins by a lot
The numbers are hard to argue with. A false story was 70 percent more likely to be retweeted than a true one. To reach 1,500 people, the truth needed about six times as long. False stories also went deeper: where a true story usually made a splash and sank, falsehoods kept hopping from person to person to person, reaching ten levels of resharing about twenty times faster than facts did. The pattern held across every topic the team looked at, and it was strongest for political news.
Everyone blamed the bots
When the study came out, the tempting conclusion was that automated accounts were doing the dirty work. Machines flooding timelines with rubbish: annoying, but at least it lets the rest of us off the hook. So the researchers re-ran their analysis with all the bots removed. The gap did not close. False news still spread further and faster, which means the sharing was done by people. Ordinary accounts, ordinary thumbs.
Take the bots out of the picture and false news still wins the race. The engine is us.
The novelty trap
Why would people be better couriers for falsehood than for truth? The team's best answer is novelty. When they compared stories to what each user had seen in the previous two months, the false ones stood out as measurably newer, stranger, more surprising. And surprise is precisely what we love to pass along. Telling someone a thing they already know is conversation; telling them a thing that makes their eyebrows jump is a gift.
The replies told the same story. False news drew reactions of surprise and disgust. True news drew sadness, anticipation, trust: quieter feelings, the kind that rarely make anyone reach for the share button. Reality, most days, is simply less exciting than fiction, because fiction is written to be exciting and reality is under no such obligation.
One platform, one decade, one study? Fair caution. But notice what the mechanism runs on. Nothing in the novelty trap is specific to one app or one era: it lives in the reader, not in the software. Feeds come and go, logos change, and the same old lever keeps working, because the lever is a preference for surprise that we all carry around. That is also why the finding has aged so well since 2018.
What to do with this
We find one detail of this study strangely reassuring: the problem is not a shadowy machine army, it's a very human weakness for the surprising. Weaknesses like that respond to training. The moment a story makes you gasp and reach for the share button is exactly the moment to slow down, because that gasp is what the story was engineered to produce. Our little team builds practice cases around that pause for a living, and if the MIT study proves anything, it's that the pause needs practising.
So next time something lands in your feed and feels too astonishing to keep to yourself, enjoy the eyebrow jump. Then ask where the story has been checked before passing it on. The truth is slower. You don't have to be.