AI-translated Hitler speeches spread on TikTok with millions of views
TikTok users have used AI translation tools to post Adolf Hitler's speeches on the platform, where some videos received over 1 million views and generated supportive comments, reportedly violating TikTok's policies against promoting hateful ideologies. The platform subsequently removed most identified videos but academics have raised concerns about algorithmic amplification of extremist content.
What happened
Videos of Adolf Hitler's speeches translated using AI have been posted on TikTok, with some garnering over 1 million views. The videos feature Hitler's speeches with captions such as 'Just listen' and are accompanied by instrumental music, including Drift Phonk tracks posted without permission.
Comments on the videos praised Hitler, with users writing 'he was a hero', 'maybe he is not a villain', 'AH was a good and kind man', and stating the videos 'changed my views on him'. Between September 2 and 3, Sky News found more than 50,000 TikTok posts using speeches from Hitler and Joseph Goebbels.
Reported impact
The spread of AI-translated Hitler speeches on TikTok violated the platform's community guidelines banning promotion of individuals who cause mass violence or promote hateful ideologies. Ian Garner, Professor of Totalitarian Studies at the Pilecki Institute in Warsaw, characterized the videos as part of a wave normalizing extremism online and described them as a 'worrying phenomenon' that draws users into 'rabbit holes' of increasingly extreme content.
Index analysis
TikTok's moderation systems failed to prevent the widespread distribution of hateful content, despite the platform's policy against promoting genocidal figures. The platform's algorithm, according to Garner, can intensify exposure to extremist content, pulling users progressively into more radical material.
Governance controls
TikTok stated it removed almost all videos identified in the Media Matters report for violating community guidelines, along with more than 1,300 pieces of related content removed proactively, and claimed to remove more than 90 percent of hateful content before it is reported. However, some videos remained online following the Media Matters report.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.