It's been just over a month since Google announced Veo 3, a new tool for generating videos using AI. In a GREGE article at the time of the tool's launch, Allison Johnson raised an alarm about the meaning of Veo 3. “[W]Johnson writes. She also called the video “hell-like realistic.”
And now it appears that a scary racist has found this tool and is using it to make a terrible racist video.
This is a major point from a new American report from Media Watchdog Group Media Matters. According to the report, users use VEO 3 to feature “videos of protesters driving in their cars,” and “videos recreating the historical trauma of marginalized groups,” as well as videos depicting racist stereotypical traffic. Media issues state that they contain tags or descriptions that specify that they were created using VEO 3.
Media Matters reports are featured in some of the images featured in these videos. In short, it's really awful and incredibly racist. What's worse is that people seem to tune them. Many of the videos cited in the report had hundreds of thousands of views about Tiktok, while some had over a million views.
As Ryan Whitwam pointed out on Ars Technica, what's worse is that videos like this are not permitted under Tiktok's Terms of Use. “We do not allow hate speech, hateful behavior, or hateful ideology,” the company said on its Community Guidelines page. “This includes explicit or implicit content that attacks protected groups.”
That leads to another uneasy thinking: if hateful content is a clear violation of the regulations on a particular platform, yet still spread to hundreds of thousands of people, what can be done to enable enforcement on this scale?
This article was featured in InsideHook Newsletter. Sign up now.
