The Persistence of Spanish-Language Misinformation on Social Media Platforms
Despite repeated warnings from US lawmakers and advocacy groups, major social media platforms like Facebook, Twitter, and YouTube are still struggling to effectively combat the spread of misinformation in Spanish. This failure has significant implications, particularly with the midterm elections approaching, as Latino voters represent a substantial portion of the electorate and rely on these platforms for information. The lack of adequate moderation in Spanish essentially disenfranchises these voters, hindering their access to accurate information and potentially influencing their electoral choices. This issue was raised over a year ago, but meaningful progress remains elusive.
The types of misinformation circulating in Spanish often mirror English-language narratives, primarily focusing on politics and health, particularly vaccines. One of the most concerning narratives is the "big lie," the false claim that the 2020 presidential election was stolen. This pervasive falsehood, widely believed within certain demographics, erodes trust in democratic institutions and discourages voter participation. The concern is that pre-emptive doubt regarding the legitimacy of the midterm elections is being sown, potentially impacting voter turnout and the overall integrity of the electoral process.
Despite platform policies prohibiting such content, an August report from Media Matters for America revealed numerous Spanish-language videos promoting the "big lie" still circulating on these platforms. The report documented various examples of baseless claims, including allegations of voting machine glitches, fake ballots, and deceased individuals voting in significant numbers. Specifically, the report identified three Spanish-language YouTube channels with a combined subscriber base of over 880,000 that have repeatedly violated content policies but remain active. This demonstrates a clear lapse in enforcement and a failure to hold accounts spreading harmful misinformation accountable.
Civil rights groups, including Free Press, have been urging tech companies to address misinformation in multiple languages for months, but their efforts have yielded limited results. Despite attempts to engage with platforms like Meta, TikTok, and Google regarding Spanish-language misinformation, these companies have not taken sufficient action. Requests for information on YouTube’s policing of Spanish election misinformation have gone unanswered, and calls for Facebook to ban stolen election theories have been ignored. This inaction allows harmful narratives to persist and proliferate, exacerbating the problem.
The continued prevalence of misinformation in Spanish is partly attributed to a lack of investment in language-specific moderation resources. Platforms have been criticized for failing to hire enough human moderators fluent in Spanish and for inadequately training artificial intelligence systems to detect misinformation in the language. This disparity in resource allocation was highlighted by whistleblower Frances Haugen, who revealed that Facebook’s misinformation spending heavily favored English content, despite a much smaller percentage of English-speaking users. This prioritization of profitability over user safety contributes to the proliferation of harmful content in non-English languages.
While social media companies claim to be taking steps to address misinformation in Spanish, including improvements to detection models and partnerships with fact-checking organizations, the impact of these efforts remains questionable. Meta, Facebook’s parent company, asserts its commitment to combating Spanish misinformation and claims to have improved its prediction models. The company cites its collaboration with 90 fact-checking organizations globally and investments in media literacy initiatives. Similarly, YouTube and Twitter emphasize their global misinformation policies and collaborations with various organizations to promote voter registration and accurate information. However, critics argue that these efforts are insufficient and lack the urgency needed to effectively combat the problem, especially given the proximity of the midterm elections. Further, the lack of transparency regarding the number of Spanish-speaking moderators employed by these companies raises concerns about their actual capacity to address the issue effectively. Ultimately, the continued presence of 2020 election misinformation on these platforms underscores the ongoing challenge and raises doubts about their commitment to truly protecting Spanish-speaking users from harmful content.