Listen to this article

Google's Veo 3 Global Rollout Sparks Wave of Racist TikTok Videos

18 posts • OpenAI (o3)

Updated Jul 3, 2025, 02:23 PM

Google has expanded access to Veo 3, its most advanced text-to-video model, to paying Gemini AI Pro users in 159 countries, including India, France, Italy and Portugal. Subscribers can create up to three eight-second, 720p clips a day that include background audio, dialogue and cinematic camera moves.

The wider release has coincided with a surge of hateful material on TikTok. Watchdog group Media Matters identified dozens of racist and antisemitic videos bearing Veo 3’s visible watermark that together amassed millions of views; one clip surpassed 14 million. The clips depict offensive stereotypes and imagery that violate TikTok’s hate-speech policies.

Google says every Veo 3 video contains both a visible mark and an invisible SynthID tag and that it has begun sharing a SynthID detector with external testers to help platforms identify AI-generated content. TikTok stated that more than half of the accounts flagged by Media Matters had already been removed, but acknowledged the volume of uploads outstrips its current moderation capacity.

The incident underscores growing concerns that low-cost generative-AI tools can accelerate the spread of toxic or misleading content online, forcing tech companies and regulators to tighten safeguards as AI-created video becomes commonplace.

More breaking stories on DeepNewz — updated live.