Google lets users drop visible AI watermarks

Google is giving users the ability to strip visible watermarks from AI-generated images, videos, and songs, a move that trades a blunt visual marker for subtler, embedded forms of provenance.

The change, announced Friday, makes the visible mark optional across several of Google's generation models. It does not touch the underlying invisible SynthID watermark or the C2PA standard-related metadata, which remain in place for transparency.

Optional marks, persistent metadata

Josh Woodward, Google's VP for Gemini, said in a post on X that the toggle will be available for Nano Banana, Omni, and Lyria models. The setting to turn the visible watermark off will appear in Gemini and Google's video editor, Flow, with support for Search coming soon.

The feature is rolling out in the coming days. Once available, users can navigate to Settings > Media Watermark to flip the visible mark on or off.

Balancing creativity and safety

The shift reflects an evolving approach to labeling AI-generated media. Visible watermarks can often make content less useful for professional and creative work, but the need to identify AI-generated content remains.

“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. So you can still use Gemini or Search to see if an image was AI-generated,”

Woodward wrote in his post. That balance means users can still check whether an image was made by AI through Gemini or Search, even when the visible mark is removed.

Watermarking evolution

The announcement comes after Anthropic’s hotly debated move to include a watermark in text and files generated by Claude to comply with EU regulations. Google's approach appears to be the opposite: rather than forcing visible marks on everything, it's making them optional while keeping the invisible checks.

Google is also open sourcing a new library called Credentio to let developers embed a local validation mechanism in their apps.

Implications for creators and consumers

For creators, the ability to remove visible watermarks could make AI-generated content more usable in professional settings, where a prominent mark might interfere with design or presentation. For consumers, the shift underscores the importance of invisible watermarking and metadata, which now carry the primary burden of provenance.

Google's move to make visible watermarks optional—while maintaining invisible ones—could become a template for how other companies navigate this space.