In 2018, during the last midterm elections in the United States, TikTok was barely beginning its global conquest and had not yet become the ultra-popular platform with more than one and a half billion active users per year. month. Four years later, on the threshold of these new midterms, the application of the Chinese giant Bytedance is this time in the crosshairs of the observers of the poll: beyond the questions on the potential access of Beijing to the data of foreign users, it is the risks of a wave of disinformation that worry, in a country where “fake news” has become a major subject of presidential campaigns since the one that brought Donald Trump to the White House in 2016.
Various articles and studies published since September have fueled these fears. An investigation by the NGO Global Witness, conducted jointly with the “Cybersecurity for Democracy” team of the Tandon School of Engineering at New York University, concludes, for example, that the video platform “fails to detect false electoral information”. Investigators say they submitted “ten advertisements in English and ten in Spanish” on different social networks, “five containing false electoral information and five aimed at delegitimizing the electoral process”. TikTok, which prohibits the purchase of political advertisements and claims to have reinforced the control of paid political content published by influencers, has let 90% of these false advertisements pass, assures their report.
The difficult moderation of videos
TikTok is not the prerogative of electoral disinformation in the United States – the same study also points to the flaws of Facebook with its 226 million American users. Still, the analysis of the video format, more time-consuming than that of the written word, complicates the fight against false information, as shown again recently the example of YouTube. A problem that is all the more common as the Chinese platform’s filmed and viral short pellets have greatly inspired its competitors, such as YouTube Shorts and Instagram Reels.
In a recent report, the Institute for Strategic Dialogue attests to having effortlessly found “content that violates the policies in place on YouTube, Instagram and TikTok against electoral misinformation”. An ease that suggests to analysts “that the new features (…) such as short videos are deployed without full risk assessment and user safety considerations”but also that moderation policies “continue to fail despite promises”.
You have 51.35% of this article left to read. The following is for subscribers only.