Google Introduces Visible Watermark in AI-Generated Videos with Veo 3

Google has introduced a visible watermark in AI-generated videos produced by its Veo 3 model. This initiative is part of a comprehensive strategy to tackle rising concerns regarding deepfakes and misinformation. The watermark is being implemented as Veo 3 becomes accessible to a broader audience worldwide, following its recent launch in the UK and 72 additional countries.

Details About the Watermark

The visible watermark is subtle and can easily be overlooked. It appears in the lower right corner of videos created through Google’s Veo 3 platform. However, videos produced via the Flow tool by users subscribed to the $249.99/month Gemini AI Ultra plan do not feature this watermark. This change was discreetly announced in a thread by Josh Woodward, Vice President at Google Labs and Google Gemini, and is further explained in a blog post that highlights the expansion of Veo 3.

In a statement, Woodward noted that visible watermarks have been added to all videos, excluding those from Ultra members in Flow. This move represents an initial step as Google strives to enhance the accessibility of its SynthID Detector.

Global Expansion Plans

Woodward also shared that Veo 3 will soon be launched in other countries, including India. Up until now, Google had primarily depended on SynthID, an invisible digital watermark embedded in all AI-generated content, to trace the origins of synthetic media. However, while SynthID is advanced, it lacks the widespread availability and functionality for typical users, which is a significant hurdle, especially as lifelike AI-generated content becomes more ubiquitous on social platforms with little to no disclosure.

Increasing Popularity of Veo 3

Veo 3 has attracted considerable attention since its introduction at Google I/O 2025. Its capability to create cinematic, photorealistic videos complete with dialogue and audio has allowed users to produce diverse content including animal fantasy clips, influencer-style advertisements, mock news segments, and unboxing videos, effectively blurring the line between real and synthetic.

Addressing Misinformation with Transparency

With the heightened realism of Veo 3 underscoring the dangers of online misinformation, the visible watermark aims to provide a more immediate layer of transparency. Nonetheless, some critics have noted that the watermark is often faint and inconsistently visible across different scenes, urging enhancements in its size and positioning.

Despite its drawbacks, the visible watermark signifies progress as Google continues to develop and expand the availability of the SynthID Detector tool, which is currently not accessible for widespread use. As the prevalence of AI-generated video increases, such visible indicators may be vital in assisting audiences to differentiate between genuine footage and synthetic media.