{
  "id": 16800649,
  "title": "Social Media When Elections Are Close: Doing the Impossible Better",
  "url": "https://urgent.news/2026/10/27/social-media-when-elections-are-close-doing-the-impossible-better",
  "topic": "tech",
  "section": "Tech",
  "published": "2026-10-27T21:20:30.000Z",
  "source": {
    "name": "EFF Deeplinks",
    "slug": "eff-deeplinks",
    "url": "https://www.eff.org/deeplinks/2026/10/social-media-when-elections-are-close-doing-impossible-better"
  },
  "original_language": "en",
  "account": "During a pivotal early voting period on October 17, Meta faced scrutiny over the suspension of a prominent political Instagram account and allegations of boosting a candidate's visibility in Brazil’s recent presidential election. Both instances were attributed to mistakes in Meta's automated content moderation systems, which struggle to distinguish genuine engagement from bot activities. Though Meta denies deliberate interference, these events underscore the urgent need for robust reforms in social media content moderation systems to protect democratic communication.\n\nThe immediate aftermath of these incidents emphasizes the critical role of accessible information in shaping political participation and speech. Automated moderation systems, while capable of processing vast amounts of content, are fraught with issues such as bias, context misunderstanding, and disproportionate impacts on marginalized communities. For instance, in 2020, Meta's algorithms were found to delete Arabic-language content 77% of the time without deleting hate speech, a disparity that worsens in less-resourced languages.\n\nThese systemic failures highlight the challenges of relying solely on automated moderation. The 2025 report by the Center for Democracy and Technology reveals inconsistencies and biases in labeled datasets, especially in languages like Maghrebi Arabic and Kiswahili, due to a lack of native speakers and language evolution. Similarly, ChatGPT’s outputs in low-resource languages reveal significant shortcomings.\n\nHowever, the situation is not insurmountable. By adhering to the Santa Clara Principles 2.0, companies can take significant strides towards more ethical moderation practices. These principles advocate for automated systems to assist, not replace, human moderators, ensuring transparency in automation usage, regular bias audits, and enhanced personal data protection. Moreover, companies must establish clear appeal processes and promptly address user grievances with human oversight.\n\nLegislators also play a crucial role by refraining from mandating specific moderation technologies and focusing instead on safeguarding human rights and preventing legislative overreach. Public education about data usage in electoral contexts is equally vital. The ubiquity of social media platforms, which wield immense influence over political discourse, necessitates a collective effort from companies, policymakers, and civil society to uphold democratic values and ensure equitable access to information.",
  "summary": "Everyone knows, or should, that perfect (balanced, targeted, transparent, swift and swiftly appealable) content moderation at scale is impossible. That doesn’t mean powerful social platforms don’t have to try to get at least reasonably close, especially when access to good information can have a profound effect on political participation and speech. That’s why so many people, including the Meta…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}