Racist videos made with AI are going viral on TikTok | The Verge

According to a report by Media Matters, racist videos created with Google's AI video generation tool, Veo 3, are gaining significant traction on TikTok. These videos, which have collectively amassed millions of views, are filled with racist tropes and target various groups, particularly Black people. The organization identified the videos as potentially generated by Veo 3 based on the "Veo" watermark present in the clips, along with related hashtags and user mentions. The videos also adhere to Veo 3's eight-second clip limit, further supporting the claim of their origin.
Google launched Veo 3 in May, allowing users to generate video clips and audio from text prompts. While Google states it will block harmful requests, the proliferation of racist content on TikTok raises concerns about the effectiveness of these safeguards. TikTok has its own policies against hate speech and hateful behavior, and a spokesperson confirmed that the platform had removed the accounts identified in the report, many of which were already banned.
The Verge also found some of the problematic videos on YouTube, though with fewer views, and Wired discovered similar content on Instagram. The content wasn't limited to anti-Black racism; Media Matters found videos containing antisemitic content and racist depictions of immigrants and Asian people, highlighting the broader scope of the issue.
The emergence of these AI-generated racist videos underscores the challenges of content moderation in the age of advanced AI tools. As AI video generation becomes more accessible, platforms like TikTok and Google will likely face increasing pressure to refine their detection and enforcement mechanisms to prevent the spread of harmful content.