Racist videos, seemingly generated by Google’s AI video tool Veo 3, have amassed millions of views on TikTok, according to Media Matters, a nonprofit media watchdog. The organization’s investigation revealed AI-generated content containing racist tropes, frequently targeting Black individuals.
Media Matters identified the videos, one of which garnered 14.2 million views, as Veo 3 creations due to a “Veo” watermark present in the corner of the clips. Additionally, some users included hashtags, captions, or usernames referencing Veo 3 or AI. Each clip discovered by the organization was eight seconds in duration, or comprised multiple segments each lasting no more than eight seconds, consistent with Veo 3’s eight-second generation limit.
Google launched Veo 3 in May, enabling users to generate AI video clips and accompanying audio from text prompts. Google states on its website that it will “block harmful requests and results.” TikTok’s platform guidelines prohibit “hate speech and hateful behavior” and specify that the platform “will not recommend content that contains negative stereotypes about a person or group with a protected attribute.”
“We proactively enforce robust rules against hateful speech and behavior and have removed the accounts we identified in the report, many of which were already banned prior to the report publishing,” TikTok spokesperson Ariane de Selliers told The Verge.
The Verge located some of the videos highlighted by Media Matters on YouTube, although with fewer views. Wired also found similar racist AI-generated videos on Instagram. Other videos identified by Media Matters included antisemitic content and racist portrayals of immigrants and Asian individuals.
