Amplifying Antisemitism: How Recommender Algorithms Serve Harmful Content to Children
Using 10 simulated 15-year-old accounts (5,500+ videos), teens were exposed to antisemitic content within hours; TikTok pathways escalate from mainstream to neo-Nazi content while Rumble surfaced overt antisemitism from the outset.
Executive summary
AI-generatedThis summary was generated by AI from the original report to make it easier to scan and cite. It is not a substitute for the source — read the original above.
Researchers tested how recommender systems on TikTok and Rumble expose young users to antisemitic content. On TikTok, 10 simulated accounts posing as 15-year-olds with varied political interests interacted with the platform over 14 days, generating more than 5,500 recommended videos analyzed thematically for content patterns. On Rumble, researchers reviewed 4,412 videos drawn from the platform's "Editor's Picks" over six months, screening for antisemitism-related keywords and closely examining 259 potentially relevant clips.
The study found TikTok's recommender system created pathways moving users from neutral lifestyle content toward increasingly politicized and conspiratorial material, including neo-Nazi content, with harmful themes surfacing in videos, comments, stickers and sounds alike. Rumble, by contrast, surfaced overt antisemitic content — slurs, Holocaust distortion, conspiracy narratives about Jewish control — more directly and from the outset.
The authors describe the findings as evidence of urgent gaps in platform accountability and call for stronger enforcement of online safety rules to protect minors from algorithmically amplified antisemitic content.
Think this summary is wrong? Contact us.