Google Is Now Letting Users Remove Its Visible AI Watermark, But There Is a Catch

Google Is Now Letting Users Remove Its Visible AI Watermark, But There Is a Catch

For months, Google’s AI-generated images, videos, and music have come with something users could not easily escape: a visible watermark.

Now, Google is changing that.

The company has announced that users will soon be able to turn off the visible watermark on AI-generated content, giving creators a cleaner final product without completely removing the technology Google uses to identify AI-made media.

And that distinction is important.

The Watermark Is Going Away But Not Completely

Google’s new option will allow users to remove the visible watermark from content created with supported AI models, including Nano Banana, Omni and Lyria.

The setting is being introduced in Gemini and Google’s AI filmmaking tool, Flow, with support for Google Search expected later. The rollout is scheduled for the coming days. 

Users will reportedly find the option under Settings → Media Watermark, where they can choose whether the visible mark appears.

For creators, this could be a welcome change.

Imagine creating a polished marketing image, presentation visual, music track, or video with Gemini only to have a visible AI label sitting on the finished work. While that label is useful for transparency, it is not always ideal when the content is being used professionally.

Google appears to be trying to solve that problem without abandoning AI-content identification altogether.

The Invisible Watermark Stays

Here is where the story gets interesting.

Turning off the visible watermark does not mean the AI-generated content becomes completely untraceable.

Google says its invisible SynthID watermark will remain. C2PA-related metadata will also continue to provide information about the content's origin and editing history.

SynthID is designed to embed an imperceptible signal into AI-generated content. Humans cannot simply look at a file and see it, but compatible verification tools can search for the signal.

Google has been expanding this technology aggressively. In May 2026, the company said SynthID had already been used to watermark more than 100 billion images and videos and around 60,000 hours of audio.

So, the visible label may disappear, but the digital fingerprint remains.

Why Google Is Making This Change

There is a simple tension behind the decision.

Creators want freedom. Platforms want transparency.

If every AI-generated image has a visible watermark, identifying synthetic content becomes easier. But professional designers, marketers, filmmakers, and businesses may prefer their final work without an obvious AI label.

Google's approach is essentially:

Let creators remove the visible sign, while keeping the hidden verification system underneath.

That means someone could potentially receive an image without a visible Google AI watermark and still use Gemini or other supported Google tools to check whether Google's AI technology was involved.

There are also regions where visible AI labels may still be required by law, meaning the removal option will not necessarily be available everywhere.

Google Is Also Opening Up Its Verification Technology

The watermark change is not happening in isolation.

Google has also introduced Credentio, an open-source C library designed to help developers work with C2PA Content Credentials and validate content locally.

That matters because the future of AI transparency probably cannot depend on one company alone.

AI-generated images, videos and audio move across Instagram, TikTok, websites, messaging apps and countless other platforms. A system that helps people understand where content came from needs to work across that ecosystem.

Google has therefore been pushing both SynthID and C2PA as part of a broader effort to make the origin and editing history of digital content easier to verify. 

The Bigger AI News

This may look like a small settings change, but it points to a much bigger shift in the AI industry.

The question is no longer simply "Was this made by AI?"

It is becoming:

"How do we prove where this content came from without getting in the creator's way?"

Google is betting that invisible watermarks and digital content credentials can provide that answer.

So yes, the visible AI watermark may soon disappear from your Gemini creations.

But the digital fingerprint is not going anywhere.

And as AI-generated content becomes almost impossible to distinguish from human-made work at a glance, that hidden fingerprint could become more important than the watermark everyone can see.