An investigation reveals how Google’s Veo 3 tool is being used to create and spread racist content targeting Black people and immigrants on social media platforms, garnering millions of views despite ban policies.
Introduction: When creativity turns to hate
At a time when tech companies are boasting about the ability of artificial intelligence tools to unleash creativity, a dark and ugly side of this technology is emerging. The non-profit media organization Media Matters has revealed a disturbing discovery: racist videos, seemingly created using Google’s Veo 3 video generation tool, are spreading like wildfire on TikTok, racking up millions of views. These videos, filled with racist stereotypes and many targeting Black people, raise serious questions about the ability of tech companies to control their tools and prevent their use in spreading hate.
How was the issue uncovered?
Media Matters researchers noticed short video clips bearing a “Veo” watermark in the corner, the mark placed by Google’s new tool. One of these clips alone garnered 14.2 million views. In addition to the watermark, some users used hashtags or comments explicitly referring to Veo 3 or artificial intelligence. Furthermore, the length of the clips—not exceeding 8 seconds or consisting of multiple short segments—matches the maximum video length that the Veo 3 tool can currently generate.
Failure of safety policies
The stark paradox lies in the fact that this content is spreading on platforms claiming to have strict policies against it. On its website, Google states that it “will block harmful prompts and outputs.” Similarly, TikTok’s rules state that “hate speech and hateful behavior have no place on TikTok.” Nevertheless, this content managed to bypass safety filters and reach millions of users.
In response to the report, Ariane Célier, a TikTok spokesperson, said: “We proactively enforce strict rules against hate speech and hateful behavior, and we have removed the accounts identified in the report, many of which were already banned prior to the report’s publication.” Despite this response, the fact that these videos reached such a massive number of views before being banned shows serious flaws in moderation systems.
A problem not limited to TikTok
TikTok was not the only platform suffering from this issue. The Verge report found some of the same videos on YouTube, albeit with fewer views. Wired magazine also discovered similar racist AI-generated videos on Instagram. In addition to content targeting Black people, Media Matters found other clips containing anti-Semitic content and racist portrayals of immigrants and Asians, confirming that the problem is widespread.
The bigger challenge: A race between creation and moderation
This incident highlights the enormous challenge facing platforms in the era of AI-generated content. While these tools make it easier for anyone to create visually engaging content, they also make it easier for malicious actors to produce massive amounts of harmful content at a pace that exceeds the ability of human moderators and algorithms to review and remove. It is an ongoing battle between the speed of creation and the effectiveness of moderation.
Conclusion: Responsibility falls on everyone
The spread of racist videos created using Veo 3 is a painful reminder that artificial intelligence tools are a double-edged sword. While they promise innovation, they carry the risk of amplifying the worst biases present in our societies. Responsibility now lies with tech companies like Google and TikTok to develop more robust and effective security systems—not just to delete harmful content after it spreads, but to prevent its creation and distribution in the first place. The future of a healthy and safe digital space depends on their success in this endeavor.
