Recipes for hydroxychloroquine made from grapefruit and lemon peel, drinks for abortion, Ukrainian soldiers who manufacture fake corpses, or even proof of the rigging of the 2020 American elections… false information abounds on TikTok. An investigation by NewsGuard — an entity founded by journalists with the aim of combating disinformation — looked into the most popular social network of the moment.
With a total of 3.5 billion downloads at the start of 2020, the application has become essential. In 2021, TikTok even overtook Google as the most popular site in the world according to Cloudfare, and was even called the “New Google” by the New York Times. What’s more, according to information from TechCrunch, 40% of Generation Z prefer to use TikTok or Instagram rather than Google for information. So, with such a large but also so young audience, ensuring the veracity of the information is essential.
However, despite the rules of use and the content removal policy brandished by TikTok, the NewsGuard study shows that approximately one in five videos contains misinformation. Using blank counts, researchers selected the top 20 results from 27 different news searches. From Covid vaccines to abortion to the war in Ukraine to the US election, each of the most popular TikToks has been analyzed. And of these 540 results, 105 videos”contained false or misleading claims“, we read in the article, or 19.4% of them.
Dangerous misinformation
If the share of videos containing false information is quite substantial, it is also the content and the dangerousness of the words relayed that are pointed out. While typing “mRNA vaccine” in the search bar, for example, researchers came across five videos promoting false claims about the Covid-19 vaccine. The videos of Doctor Robert Malone, known for the misinformation he spreads around the vaccine, were even recurrent. Same (sad) observation about abortion: NewsGuard had already revealed that videos about dangerous and unapproved abortion techniques had multiplied since a highly contested decision by the United States Supreme Court.
If TikTok moderated the content by displaying a message explaining that the search “mugwort abortion” (Editor’s note, a plant) “may be associated with behavior or content that violates our rules”, it is still possible to circumvent the system by asking the question: “Does mugwort cause abortion?”. In this specific case, 13 of the top 20 results advise abortion methods presented as natural, but without scientific basis, herbal.
A very unreliable search engine
The major problem noted throughout the survey is not so much the presence of these videos as their highlighting and referencing. Each of the searches carried out on the Chinese social network was also carried out on Google and, each time, the engine did not present any misinformation in the first results, and even directed the searches. For example, typing “mugwort abortion” on Google, scientific articles exposing the dangers of this type of practice stand out. The same goes for the word suggestion in the search bar.
According to NewsGuard’s examples, “when a user types in the term “climate change”, TikTok suggests searching for “climate change demystified” or “climate change doesn’t exist”. To a user searching for “covid vaccine”, TikTok offers to search for “covid injury vaccine”, “covid vaccine truths”, “covid vaccine unmasked”, “covid hiv vaccine”, and “covid vaccine warning””. The result is very different with Google: “Suggested terms in Google searches are more direct. For example, for the search “covid vaccine”, Google suggests “covid vaccine without appointment”, “which covid vaccine is the best” and “type of covid vaccine”. None of these terms were suggested on TikTok.”
For its part, the Chinese firm declares on its site that “TikTok content is first subjected to technology that identifies and flags potential rule violations, such as adult nudity and violent and graphic content. When content is identified as violating our Community Guidelines, it is automatically removed or flagged for further review by our security team.“A moderation which is therefore primarily based on artificial intelligence. Contacted by NewsGuard about the misinformation on the platform, TikTok gave no response. Either way, these discoveries come at a bad time, as they emerge as the social network seeks to give itself an educational and learning-oriented image by launching the #TikTokTaughtMe (“TikTok taught me”). .