Gemini Lets You Turn Off Visible AI Watermarks

If you have ever generated a product image, a social media graphic, or a quick promotional video with Google’s Gemini and gotten annoyed by that little sparkle logo stamped in the corner, you are not alone. Enough people complained about it that Google actually did something about it.

On August 14, 2026, Google confirmed a new setting that lets you switch off the visible AI watermark on content created with Gemini. This covers images made with Nano Banana, videos made with Omni, and music made with Lyria, with Search support said to be coming soon. It is a small toggle, but it solves an annoying problem for anyone using Gemini to create client work, marketing assets, or polished business content.

But before you go flipping that switch and assuming your AI-generated content is now indistinguishable from anything else, there is an important catch. The visible watermark disappearing does not mean the content is untraceable. Google is keeping an invisible layer, called SynthID, along with C2PA provenance metadata, embedded in every file regardless of what you choose in settings.

This update matters more than it looks at first glance. Here is what changed, what stays hidden behind the scenes, and how to use it without running into trust or compliance issues.

What Exactly Changed With the Visible Watermark in Gemini

Here is the update itself, based on Google’s own announcement and the reporting that followed.

Google’s VP for Gemini, Josh Woodward, confirmed the change directly on X, describing a new Media Watermark setting rolling out across the Gemini app and Google’s Flow video editor.

✅ Papercut fixed: You can now toggle visible watermarks on or off in Gemini and Flow, with Search coming next.

This applies to watermarks on all images (Nano Banana), videos (Omni), and songs (Lyria) except in countries where it's required by law to keep them. pic.twitter.com/utHN0yDmD3

— Josh Woodward (@joshwoodward) August 14, 2026

The rollout began August 14, 2026, and Google said it would reach accounts gradually over the following days.

Where to Find the Toggle

The setting lives under Settings, in a section labeled “Media Watermark.” From there, you can switch the visible mark on or off, and your preference applies to all future images, videos, and music tracks you generate.

What It Covers

• Nano Banana (Gemini’s image generation model)
• Omni (Gemini’s video generation model)
• Lyria (Gemini’s music generation model)

Google has also said support for this toggle is coming to Google Search’s AI creation features next, though that had not rolled out at the time of writing.

Where It Does Not Apply

A few important exceptions to know about.

• India, South Korea, and Vietnam: In these markets, the option is currently limited to Google AI Ultra subscribers. Free and standard plan users in these countries keep the visible watermark by default.
• Work and school accounts: Google Workspace accounts do not get this setting at all. Visible watermarks stay mandatory regardless of preference.
• Regions with legal labeling requirements: In countries where local law requires visible AI labeling on generated content, Google may restrict or override the toggle entirely.

Quick Reference Table

What Does Not Change: SynthID and C2PA Metadata

This is the part that matters most, and the detail a lot of quick news coverage glosses over. Turning off the visible watermark only removes what you can see. It does not remove the invisible layer of tracking Google has built into every single generation.

What Is SynthID?

SynthID is Google DeepMind’s imperceptible watermarking technology. It is built into the content itself at the pixel, frame, or audio level, depending on the media type, and it is designed to survive things like screenshots, cropping, resizing, and format conversion. Google has stated that SynthID has now been applied to more than 10 billion pieces of content across its products.

There is no user-facing way to disable SynthID. It is embedded automatically, every time, regardless of your visible watermark setting.

What Is C2PA Metadata?

C2PA stands for the Coalition for Content Provenance and Authenticity, an industry standard used across parts of the tech and media world to record information about how a piece of digital content was created. Every file Gemini generates carries this metadata, which records its AI origin.

Why This Matters

Together, these two systems mean that even with the visible watermark switched off, Gemini and Google Search can still confirm whether a piece of content was AI-generated. The transparency has not gone away. It has just moved from something visible on the surface to something detectable underneath.

This is an important distinction for business owners to understand before using this feature for anything client-facing or brand-sensitive. A cleaner-looking image does not mean an untraceable one.

Why Google Made This Change

According to Woodward, Google gathered direct feedback from users about their biggest frustrations with Gemini, and removing Nano Banana’s visible watermark landed in the top 10 requests out of thousands of responses.

The reasoning makes sense from a practical standpoint. A visible watermark is fine on a quick social media post, even useful there. But it becomes a real problem on a client presentation deck, a printed poster, a product mockup, or polished marketing material, where a stray logo in the corner reads as sloppy rather than transparent.

Google’s own framing, in Woodward’s words, was about striking “a balance between creative control and safety.” Businesses get cleaner, more professional-looking output for legitimate creative use, while Google keeps its underlying detection and provenance systems fully intact for accountability purposes.

There is also a competitive angle. Nano Banana competes directly with a growing field of free AI image generators, many of which do not apply visible watermarks at all. A cleaner default output makes Gemini’s content easier to use across marketing and creative workflows, closing a real usability gap against competitors.

Is the Visible Watermark Toggle Good or Bad for Business?

Like most platform changes, the honest answer depends on how you plan to use it.

The Upside

• Cleaner, more professional output. For business owners using Gemini to create marketing graphics, blog images, product mockups, or presentation visuals, removing the watermark means content that looks finished and polished without extra editing steps. For anyone optimizing AI-generated visuals for search, a complete image SEO audit ensures those visuals also perform well in search results.
• Faster workflows. No more cropping out corner logos or running images through separate watermark removal tools before publishing.
• Better client-facing deliverables. Agencies and freelancers using Gemini for client work can now deliver assets that do not visibly announce their AI origin, which matters for professional presentation.

The Catch

• Traceability has not disappeared. Because SynthID and C2PA metadata remain embedded, content can still be identified as AI-generated by detection tools, platforms, and increasingly by Google Search itself. If your brand values full transparency, or operates in a regulated industry, this is worth factoring into your content policy.
• Trust and disclosure considerations. Just because you can hide the visible mark does not mean you should skip disclosing AI use where it matters, particularly for testimonials, before-and-after imagery, or anything that could mislead a customer if presented as authentic, non-AI content.
• Platform and regional inconsistency. Since the rollout varies by account type and region, check settings across your team’s accounts to make sure the watermark preference matches what each piece of content needs.

Practical Tips for Using This Responsibly

Here is how business owners can make the most of this update without running into trust or compliance issues.

1. Use It for Legitimate Creative Polish, Not Deception

Turning off the watermark is great for internal decks, blog graphics, and general marketing visuals. It is not a tool for making AI content pass as something it is not in contexts where authenticity matters, like customer testimonials or real product photography.

2. Keep a Disclosure Practice for Sensitive Content

Even with the visible mark gone, consider adding your own disclosure language where appropriate, especially for content touching on health, finance, legal topics, or anything where AI generation should be clear to your audience.

3. Check Settings Across Your Team

If multiple people on your team generate content through Gemini, do not assume everyone has the same watermark preference set. Standardize your team’s approach based on what each piece of content is actually used for.

4. Understand That Detection Tools Still Work

Do not rely on the absence of a visible watermark as proof your content cannot be identified as AI-generated. SynthID and C2PA metadata mean detection remains possible, and increasingly, Google Search itself may start surfacing this information.

5. Watch for the Search Rollout

Since Google has said this feature is coming to Search’s AI creation tools next, keep an eye on how that expands, especially if you use Google’s AI features directly within Search for content creation.

6. Combine AI-Generated Visuals With Original Brand Assets

For the strongest, most trustworthy brand presence, mix AI-generated visuals with real photography, original graphics, and authentic behind-the-scenes content. This builds customer trust that a clean AI image alone cannot fully replicate.

What This Means for Local Businesses and Startups

If you are a local business or a lean startup team using Gemini to speed up content creation on a tight budget, this update is mostly a practical win. Product mockups, social graphics, seasonal promotional content, and quick marketing visuals will now look more polished straight out of the tool, without extra editing steps.

That said, local businesses in particular tend to rely heavily on trust and authenticity, real photos of your storefront, your team, your actual product. AI-generated visuals are useful for supplementary content, promotional graphics, and creative concepts, but they should not fully replace original brand imagery, especially on your Google Business Profile and core website pages, where authenticity tends to perform better with both customers and local search algorithms. For a full breakdown of building visibility across both traditional search and AI tools, see our steps to make your business site visible to AI.

How the F9XR Team Can Help

Features like this Gemini watermark toggle can change how your business creates and presents content. The F9XR Team helps business owners, startups, and local businesses build a content and digital strategy that uses AI tools without sacrificing trust or brand authenticity. That includes:

• Guidance on responsibly integrating AI-generated visuals into your marketing and website content
• Website development and website redesign work that balances polished AI-assisted visuals with authentic, original brand imagery
• Local SEO strategy that keeps your Google Business Profile and website content trustworthy, consistent, and backed by real brand assets
• Content policy support to help your team decide when AI-generated content needs disclosure and when it does not
• Ongoing digital presence management so you are never caught off guard by fast-moving platform updates like this one

If you are unsure how to fit new AI creative tools into your existing brand strategy without losing customer trust, reach out and we can work through it together.

Key Takeaways

• Google added a Media Watermark toggle in Gemini and Flow, letting users switch off the visible AI watermark on images (Nano Banana), video (Omni), and music (Lyria), announced August 14, 2026.
• Search support for this toggle is coming soon, though not live yet at the time of this article.
• The invisible SynthID watermark and C2PA provenance metadata remain embedded in every generation regardless of the visible watermark setting, and there is no way to disable either.
• The toggle is not available to everyone. India, South Korea, and Vietnam currently limit it to AI Ultra subscribers, and Google Workspace accounts do not get the option at all.
• This update is useful for cleaner, more professional-looking marketing content, but it does not change the fact that AI-generated content remains detectable behind the scenes.
• Business owners should keep disclosure practices in place for sensitive content and continue blending AI visuals with authentic, original brand imagery.

Conclusion

Google letting you turn off the visible watermark in Gemini is a welcome, practical update for anyone using Nano Banana, Omni, or Lyria to create marketing content, client deliverables, or polished business visuals. It solves a real annoyance without actually removing accountability, since SynthID and C2PA metadata keep every file traceable behind the scenes.

The smart move for business owners is treating this as a creative convenience, not a shortcut around transparency. Use it to make your content look more professional, keep disclosure practices in place where they genuinely matter, and continue balancing AI-generated visuals with authentic brand imagery your customers can actually trust.

If you want help building a content strategy that uses tools like Gemini wisely while keeping your brand’s website, local SEO, and digital presence consistent and trustworthy, the F9XR Team works with business owners and local brands on exactly this kind of website development, website redesign, and digital presence strategy every day.