According to findings on the issue of non-profit media watchdog media, racist videos that appear to be created with Google’s AI video generation tool Veo 3 have been raked up in millions of views across Tiktok. The AI-generated videos revealed by the organization are filled with racist ratios, many of which target black people.
The media issue is that because of the “VEO” watermark in the corner of the clip, Google Veo 3 determined that it had generated a video that received 112.2 million views. According to media issues, some users have listed hashtags, captions, or usernames related to VEO 3 or AI. Each clip found by an organization lasts just 8 seconds, or “each consisted of multiple clips within 8 seconds” is tracked with the 8-second limit on Veo 3.
VEO 3, launched by Google in May, allows you to generate AI video clips and audio with just a text prompt. Google says on its website it “blocks harmful requests and consequences.” Tiktok’s rules similarly state that “they don’t exist in Tiktok for hate speech and hate behavior,” and that the platform “does not recommend content that contains negative stereotypes about people or groups with protected attributes.”
“We actively implemented robust rules for hateful speeches and behaviour, and deleted the accounts we identified in the report, many of which were already banned prior to the publication of the report,” Tiktok spokesman Arian de Serie said in a statement in The Verge.
The Verge also has some of the videos highlighted in media issues on YouTube, but with fewer views, Wired also discovered videos generated on Instagram that were similarly racist AI. Other videos found in media issues include anti-Semitic and racist portrayals of immigrants and Asians.
Google did not immediately respond to Verge’s request for comment.
Updated July 2: Tiktok’s statement has been added.
