SAN FRANCISCO — In the weeks leading up to the U.S. presidential election, TikTok approved ads that included election misinformation, despite maintaining a prohibition on political advertisements, as reported by the nonprofit organization Global Witness.
Global Witness conducted an experiment by submitting eight deliberately misleading ads to assess the effectiveness of content moderation systems used by various social media platforms. This follows a similar investigation they carried out two years ago, which revealed improvements in companies’ ability to handle misleading content, particularly on Facebook. However, Global Witness highlighted TikTok’s failure, noting that the platform allowed four out of the eight submissions containing inaccuracies regarding the election to be approved. This occurred despite TikTok’s longstanding ban on political ads, instituted in 2019.
It’s important to note that these misleading ads never went live on TikTok, as the organization withdrew them before they could be published. “Four ads were incorrectly approved during the first stage of moderation, but did not run on our platform,” explained TikTok representative Ben Rathe. He reaffirmed the platform’s commitment to its advertising policy regarding political content.
In contrast, Facebook, which operates under Meta Platforms Inc., demonstrated better performance in this regard, approving only one of the eight ads submitted by Global Witness. In response to the findings, Meta stated that while the report’s scope is limited and not representative of their overall policy enforcement, they remain dedicated to continuously refining their strategies for managing content.
YouTube, owned by Google, was found to have the most stringent moderation among the platforms, approving four ads, but ultimately preventing their publication. The service requested additional identification from the testers before allowing the ads to go live and paused the account when the required information was not provided. The report raises questions regarding whether the ads would have faced approval if Global Witness had complied with the identification request. Google did not provide an immediate response when contacted for feedback.
It is common for social media companies to enforce more rigorous policies for paid advertisements compared to standard user-generated posts. The ads utilized in Global Witness’s examination included outright false information about voting processes, such as the misleading claim that Americans could cast their votes online. Additionally, other ads perpetuated false narratives that aimed to discourage voter participation, including false assertions that potential voters needed to pass an English proficiency test. Some ads even incited violence or threatened electoral personnel and procedures.