Research: social media accounts pushing misinformation are targeting users in non-English languages before the US midterms, with scant resistance from platforms
Tiffany Hsu / New York Times : Tweets: @hispaniccaucus , @grahamallen_1 , @themexican , @davidlaz , @meteor_blades , @tiffkhsu , @niubi , @researchbuzz , and @ceciliakang Tweets: @hispaniccaucus : “This year, the Congressional Hispanic Caucus pushed Meta, TikTok, YouTube and Twitter for meetings with their top executives to discuss the spread of misinformation in Spanish.” https://www.nytimes.com/... Graham Allen / @grahamallen_1 : If Alex Jones has to pay $1B how much should Facebook, Twitter, and all media outlets pay for suppressing the Hunter Biden laptop story?! Raul Gutierrez / @themexican : I'm glad the @nytimes is covering part of this story ( https://www.nytimes.com/...), but they don't mention that much of the extreme content would be flagged if it were in English. The attitude of the networks is that they don't have the resources to deal w/ it. David Lazarus / @davidlaz : Misinformation Swirls in Non-English Languages Ahead of Midterms https://www.nytimes.com/... @meteor_blades : The story is good but “misinformation” happens when unintentional mistakes are made. When the effort to persuade and confuse is based on lies, it's “disinformation,” and you can be sure the writer and editor know the difference, which isn't new. https://www.nytimes.com/... Tiffany Hsu / @tiffkhsu : More multilingual fact checking outfits ahead of the midterms are dealing with misinfo in more languages on more topics and across more platforms. “We feel like we are fighting a giant,” @TamoaC said. “We are doing our work, but Big Tech can do more.” https://www.nytimes.com/... Bill Bishop / @niubi : Misinformation Swirls in Non-English Languages Ahead of Midterms https://www.nytimes.com/... Misinformation swirling in Chinese on Twitter, YouTube and WeChat about mail-in ballots, school curriculums and hate crimes “has dangerous implications” this year for Asian American voters Tara Calishain / @researchbuzz : This has been a longstanding problem and I'm glad to see mainstream media talking about it more. https://rbfirehose.com/... https://twitter.com/... Cecilia Kang / @ceciliakang : Non-English Misinformation in the Midterms: They are in more languages On more topics Across more digital platforms With scant resistance from social media companies by @tiffkhsu https://www.nytimes.com/...
Context & Ripple Effects
This is a repeat finding, not a new one: after the 2020 vote, Facebook and Twitter were already reported to have overlooked a deluge of Spanish-language disinformation aimed at US voters, and the same gap is now documented ahead of the midterms. The Congressional Hispanic Caucus has pressed Meta, TikTok, YouTube and Twitter for meetings with top executives over Spanish-language misinformation, with little to show for it.
The coverage has also widened beyond Spanish: WeChat is flagged as a main venue for Chinese-language misinformation reaching Asian American voters on ballots, curricula and hate crimes. Meanwhile the one documented success in this corpus — YouTube cutting recommendations to fringe channels — shows enforcement works where it is applied, which sharpens the question of why non-English feeds get so little of it.
First-order effects
- Spanish- and Chinese-speaking voters head into the midterms as the least-moderated audience on the major platforms, with the Congressional Hispanic Caucus's requests for executive meetings at Meta, TikTok, YouTube and Twitter going unmet.
Second-order effects
- Advocacy groups and researchers are forced to act as the de facto moderation layer — documenting and publicizing non-English misinformation themselves — shifting the burden from platforms to civil society and ethnic media.
Third-order effects
- If the English-first enforcement pattern holds across a third consecutive election cycle, non-English moderation becomes a standing regulatory and legislative talking point, since platforms have demonstrated they can suppress fringe content when they choose to.
The trend: Platform content moderation remains structurally English-centric, pushing misinformation toward Spanish-, Chinese-, and other non-English-speaking communities that platforms consistently under-enforce.