Most fact-checking arrives after the damage is done. A false claim goes viral. Days later, someone publishes a correction that only a small slice of the audience ever sees. * Prebunking tries to flip the order. Instead of chasing each lie, you warn people about the trick before they see it. * Here is the short answer. Prebunking means teaching people the common tricks used to mislead them — like blaming one group for a big problem, or pretending there are only two choices — before they run into those tricks online. It does work, a little. Big real-world tests show small gains. And some researchers say part of the gain is people simply trusting everything less. * ## The idea is older than the internet * In the early 1960s, American psychologist William McGuire ran a series of experiments with a colleague, Demetrios Papageorgis. McGuire had been bothered by stories that US soldiers captured in the Korean War were "brainwashed". His view was that they had simply never been prepared for the arguments they would face. * His idea came from medicine. A vaccine gives your body a weak version of a virus, so it learns to fight the real thing. McGuire thought a weak version of a bad argument, plus an explanation of why it is wrong, could protect a person's mind in the same way. He called it inoculation theory. * ## From a theory to a game and a YouTube ad * About 60 years later, researchers at the University of Cambridge, led by Sander van der Linden and Jon Roozenbeek, turned the idea into tools. * The first was an online game called Bad News. You play the villain: you run a fake news site and learn the tricks by using them. In a 2019 study of about 15,000 players, people rated fake headlines as less reliable after just 15 minutes of play. The game has since been played by more than a million people, in over 20 languages. * Then they tried ads. Working with Jigsaw, a unit of Google, they made five short videos. Each one explained one trick, using cartoons and films people already know, like Family Guy and Star Wars. ("Only a Sith deals in absolutes" was their example of a false choice.) * The results came out in the journal Science Advances on August 24, 2022. The videos ran as YouTube ads in the US. About 5.4 million people saw them, and about 1 million watched for at least 30 seconds. Of those, 22,632 answered a quick test question. * On average, people who saw a video were about 5% better at spotting the trick than people who had not. Each meaningful view cost about 5 US cents. * ## Google took it to Europe * In autumn 2022, Jigsaw ran prebunking videos in Poland, the Czech Republic and Slovakia. They targeted two false stories about Ukrainian refugees: that refugees were to blame for rising prices, and that they were violent. * The videos got more than 38 million views and reached almost a third of the population in those three countries. Viewers got better at spotting the tricks and were less likely to share false claims. * But the results were uneven. The videos worked best in Poland. In Slovakia, where the videos were dubbed instead of made locally, there was "little to no discernible effect", according to Jigsaw. * In 2024, before the European Parliament elections, Jigsaw went bigger. A study published in the journal Communications Psychology on March 17, 2026 says the campaign reached more than 120 million YouTube users in 12 EU countries, including Ireland, Germany, France and Poland. * The researchers surveyed 19,735 viewers aged 45 and over. The videos helped people spot the tricks. But the authors admit the effects "were generally small". In some tests in Lithuania, the Netherlands and Romania, they found no effect at all. * ## The catch: are people learning, or just doubting everything? * A good media-literacy tool should help you tell true from false. A bad one just makes you doubt everything — including true news. * In 2023, two UK psychologists, Modirrousta-Galian and Higham, looked again at earlier data from Bad News and a similar game called Go Viral! Players did get better at calling out fake headlines. But they also started calling true headlines fake. A lot of the "improvement" was general suspicion, not sharper judgement. * A 2025 attempt to repeat an Indian study of Bad News, by the same research team, found no clear improvement at telling real headlines from fake ones. * Even the Cambridge side is careful about the hype. "You can't really expect miracles," Roozenbeek told TIME magazine in April 2024. * ## Why this matters when you read the news * Suspicion is not the same as understanding. If a tool only teaches people to shrug and say "it's all fake", the people pushing real lies have won anyway. * The useful part of prebunking is the list of tricks. They are old, and they show up in real newspapers too, not just on shady websites. Once you can name them, you start seeing them in headlines everywhere. * ## The five tricks from the YouTube study, in plain words * Scapegoating: blaming one group for a big, complicated problem. * False choice: "Either you back this, or you want disaster." As if there were no third option. * Emotional language: words picked to scare or anger you, not to inform you. * Incoherence: pushing two claims at once that cannot both be true. * Personal attacks: going after the person instead of what they actually said. * ## Related guides * Want to go further? Read our guides on the backfire effect (does correcting a lie make it worse?), whataboutism, astroturfing, and what fact-check ratings actually mean.