Google Makes Visible Watermarks Optional for AI-Generated Media

Google now lets users remove visible watermarks from AI-generated images, videos and songs while retaining SynthID and C2PA provenance signals.

Aug 15, 2026 - 13:56
 2
Google Makes Visible Watermarks Optional for AI-Generated Media
Image Credit: Chatgpt

Google is making visible watermarks optional on AI-generated images, videos, and music optional, giving users more control over how generated media appears while retaining other methods for identifying AI content.

The change applies to content created with Google’s Nano Banana, Omni and Lyria models, according to Josh Woodward, Google’s vice president for Gemini. Users will be able to turn off the visible watermark in Gemini andGoogle’ss Flow video editor, with support for Search expected later.

Google makes visible AI watermarks optional

Once the feature becomes available, users can manage it through Settings > Media Watermark and choose whether a visible mark appears on generated content. Google said the option is rolling out over the coming days.

Removing the visible watermark will not remove Google’s invisible SynthID watermark or metadata associated with the C2PA content provenance standard. Those systems are designed to preserve signals that can help identify media as AI-generated even when no visible label appears on the content.

Woodward said the change is intended to give creators more flexibility while maintaining transparency around AI-generated material.

"We're striking a balance here between creative control and safety: while the visible watermarks are now optional, invisible SynthID watermarks and C2PA metadata are still being used for transparency," Woodward wrote.

Google maintains AI verification tools

Google says users can still use Gemini or Search to check whether an image contains signals indicating that it was AI-generated. The company is therefore changing the presentation of its watermark rather than eliminating its underlying provenance measures.

Google is also open-sourcing a new library called Credentio, which is intended to help developers add local validation capabilities to their applications.

The move reflects Google’s latest approach to labelling generative AI output, allowing visible branding to be removed while keeping less-intrusive identification mechanisms attached to generated media.

What's Your Reaction?

Like Like 0
Dislike Dislike 0
Love Love 0
Funny Funny 0
Angry Angry 0
Sad Sad 0
Wow Wow 0
Shivangi Yadav Shivangi Yadav reports on startups, technology policy, and other significant technology-focused developments in India for TechAmerica.Ai. She previously worked as a research intern at ORF.