Racist AI-generated videos are all over TikTok, thanks in part to Google’s Veo 3 tool

Racist AI-generated videos are going viral on TikTok, according to a new report from Media Matters for America, a progressive watchdog group.

The videos, which appear to be created with Google’s AI video generation tool Veo 3, are filled with racist tropes targeting Black people, immigrants, Jews, and Asian people. Each video has gained hundreds, thousands, or even millions of views.

The nonprofit media watchdog identified the “Veo” watermark in the corner of the clips, along with hashtags, captions, or usernames referencing Veo 3 or AI. Veo 3, launched by Google in May, allows users to generate AI video clips and audio using only a text prompt. Videos made with the tool are limited to eight seconds. Each clip flagged by Media Matters lasted eight seconds or was composed of multiple segments that were each eight seconds or less.

Among the clips flagged by Media Matters, one labeled “Average Waffle House in Atlanta” depicted a restaurant filled with AI-generated monkeys holding buckets of fried chicken and watermelon—a scene steeped in racist imagery. That video, and others like it, amassed hundreds of thousands of views and were met with comments that echoed and amplified the stereotypes.

Additional videos uncovered by Media Matters include antisemitic tropes, reenactments of concentration camps, and racist portrayals of immigrants and Asian people. Some feature AI-generated Bigfoot characters, a social media trend that began with a humorous intent but has since been co-opted to dehumanize minority groups.

The Verge reported that some of the same videos highlighted by Media Matters were also found on YouTube, though they received significantly fewer views there. Wired found similar content on Instagram.

TikTok’s policies state that it “will not recommend content that contains negative stereotypes about a person or group with a protected attribute,” and that it uses a “combination of technology and human moderators to detect and remove accounts and content that break our rules.” Google also states it will “block harmful requests and results” with its Veo 3 feature. Fast Company has reached out to Google and TikTok for comment.

With tools like Veo 3, producing photorealistic AI videos has never been easier. The same is true for using AI-generated content to spread disinformation and target marginalized communities.

https://www.fastcompany.com/91363986/racist-ai-generated-videos-are-all-over-tiktok-exploiting-googles-veo-3-tool?partner=rss&utm_source=rss&utm_medium=feed&utm_campaign=rss+fastcompany&utm_content=rss