POLISCOPE
Back to feed
NewsThe GuardianMay 6, 2026United States

Study Finds TikTok Algorithm Prioritized Pro-Republican Content in 2024 Election

A study published in the journal Nature suggests that TikTok's recommendation algorithm systematically favored pro-Republican content over pro-Democratic content in three U.S. states during the 2024 election cycle. Researchers used hundreds of automated accounts to track feed recommendations, finding that the platform exposed users to different political narratives based on their perceived partisan leanings.

Read the full story at The Guardian

Why It Matters

The findings highlight how ic curation on major social media platforms may influence the political information environment for millions of young voters, raising questions about platform transparency and accountability in electoral processes.

Key Facts

  • Researchers used 323 dummy accounts to mimic user behavior in New York, Texas, and Georgia over 27 weeks.
  • Bots trained on pro-Republican content were 11.5% more likely to see content agreeing with their views compared to pro-Democratic bots.
  • Bots trained on pro-Democratic content were 7.5% more likely to be shown pro-Republican content.
  • Democratic-leaning accounts were shown significantly more anti-Democratic content than Republican accounts were shown anti-Republican content.
  • Pro-Democratic accounts were fed more cross-partisan content regarding immigration and crime.
  • Pro-Republican accounts were fed more cross-partisan content regarding abortion.
  • The study analyzed over 280,000 recommended videos using human and AI review.
  • TikTok disputes the study's methodology, stating it does not reflect real user behavior or the platform's customization tools.
  • The study did not analyze the actual impact of these recommendations on voter beliefs or behavior.
  • The study was limited to English-language transcripts and the early stages of a user's platform experience.

Who's Mentioned