The first major European survey on online disinformation unveils a unique map of risks on digital platforms. Conducted by the SIMODS consortium in four countries (France, Spain, Poland, and Slovakia), this study reveals that social media platforms are not all equally vulnerable to misinformation. And some even seem to reward it.
Striking differences between platforms
By analyzing users' actual exposure to public content of general interest, researchers found a worrying prevalence on TikTok, where one in five posts contains false or misleading information. Facebook follows closely with 13% of identified misinformation, closely followed by X (formerly Twitter) at 11%. YouTube and Instagram remain below 10%, but are still exposed. The only notable exception is LinkedIn, which brings up the rear with a marginal rate of 2%, indicating moderation or a network dynamic less conducive to the spread of misinformation. Beyond the sheer volume of misinformation, it is its amplification that raises concerns. The authors measured what they call the "misinformation premium": an interaction advantage systematically enjoyed by accounts deemed unreliable. Specifically, on YouTube, these accounts garner eight times more interactions per post per 1,000 subscribers than those considered trustworthy. The effect is almost as strong on Facebook and Instagram, and more moderate on TikTok. On LinkedIn, however, no significant advantage was detected, which underlines the protective effect of its professional uses.
Striking differences between platforms
Another damning finding: discredited actors don't just exist on one platform; they're everywhere. X and Facebook are their preferred hunting grounds, YouTube welcomes them on equal footing with reliable sources, while Instagram and LinkedIn are more resistant. These profiles often manage to build significant subscriber bases, sometimes approaching those of reputable sources, particularly on YouTube. Regarding monetization, data is sorely lacking. YouTube and Facebook suggest a form of indirect compensation for disinformation producers, but vague rules and a lack of transparency prevent any systematic comparison. The report calls for the activation of Section 40 of the Digital Services Act to compel platforms to disclose revenue generated per account and per piece of content. This lever is considered essential for understanding the true economics of disinformation. The conclusion is unequivocal: the very design of the platforms influences the visibility of disinformation. Where algorithms or editorial policies favor sensationalist content, fake news thrives. Conversely, rigorous technical choices can hinder this dynamic. LinkedIn, while imperfect, demonstrates this. For European authorities, these structural indicators must now guide regulation, otherwise misinformation risks becoming entrenched in the public debate.
