Google has finally responded to user complaints about the visible watermark stamped on AI-generated content in Gemini. The company is rolling out a new setting called Media Watermarks that gives users the ability to toggle visible watermarks on or off across all creations made within the Gemini app. This includes images produced by Nano Banana, videos generated with Omni, and music composed by Lyria. Previously, the visible watermark was applied automatically to every output, leaving users with no way to remove it.

What is the Media Watermarks setting?

The Media Watermarks setting is a simple on-off switch buried inside Gemini's preferences. Once enabled, users can choose whether future generations will carry a visible watermark. The setting applies globally to every image, video, and music track created through the app from that point forward. Google confirmed the rollout will happen over the next few days, so not all users may see the option immediately.

To turn off the visible watermark, users need to navigate to Gemini's settings, select Media Watermark, and toggle the option off. There is no need to re-generate previous creations; the setting only affects new outputs. Google made it clear that the invisible SynthID watermark and C2PA metadata will remain embedded in every piece of AI-generated content, regardless of the visible watermark setting. This means the origin of AI-generated content remains identifiable behind the scenes, even if users choose to hide the visible marker.

How to turn off AI watermarks for Nano Banana, Omni, and Lyria in Gemini

The steps to turn off AI-generated watermarks are straightforward and consistent across all Gemini creation tools. Users should open the Gemini app and go to the settings menu. From there, they need to locate the Media Watermark option and simply toggle it on or off. The choice applies to all future image, video, and music creations generated with Nano Banana, Omni, and Lyria.

• Open Gemini and access Settings
• Select Media Watermark
• Toggle the switch to your preferred setting

It is important to note that turning off the visible watermark does not make the content untraceable. Google still embeds an invisible SynthID watermark and C2PA metadata into everything users generate, regardless of the visible watermark setting. These invisible markers allow the origin of AI-generated content to be identified through backend verification systems.

Where is the watermark toggle available?

The rollout is not universal, and some users will not have access to the new setting. Google has restricted availability in certain regions and for specific account types. If you are located in India, South Korea, or Vietnam, you will only get this option if you have a paid AI Ultra subscription. Users in those countries without the subscription will continue to have visible watermarks automatically applied to their AI-generated content.

Work and school accounts do not get this setting at all. That means users with enterprise or educational accounts will see visible watermarks on all future Gemini creations, no matter what. Google has not provided an official reason for these restrictions, but it is likely related to regional regulations and organizational policies regarding AI transparency and accountability.

Why did Google add the AI watermark toggle to Gemini?

This update did not come out of nowhere. Google VP Josh Woodward recently asked users to share their biggest Gemini frustrations, and thousands of users responded. Removing Nano Banana's visible watermark landed in the resulting top 10 requests. Woodward said the team would consider it, and now the option has officially arrived. The decision also follows growing community efforts to strip AI watermarks from generated content altogether, with many users using third-party tools or manual editing to remove visible markers.

The introduction of the toggle represents a shift in how Google balances transparency with user experience. While visible watermarks are no longer mandatory for all users, the company insists that the underlying traceability mechanisms remain intact. SynthID is a deep watermarking technology that embeds a digital signature into pixel, audio, or video data in a way that is imperceptible to humans but detectable by specialized algorithms. C2PA metadata adds provenance information that can be verified through cryptographic signatures, helping to establish the authenticity and origin of digital content.

Google's move is seen as a response to a broader industry debate about AI-generated content and disclosure. Many platforms have faced criticism for making it too easy to generate realistic media without clear labeling. At the same time, users and creators have argued that visible watermarks can harm the aesthetic quality of generated images, videos, and music, especially when they are used for professional or creative projects. By introducing a toggle, Google is attempting to offer more flexibility while still preserving a layer of responsibility.

The company has also been under pressure from competitors who have adopted more lenient watermarking policies. Some AI image generators allow users to disable watermarks entirely with no automated tracking, while others embed hidden metadata only. Google's hybrid approach may be a way to differentiate itself as both user-friendly and accountable.

For professional creators, the ability to turn off visible watermarks could make Gemini a more viable tool for commercial projects. Photographers, graphic designers, musicians, and video producers often need clean outputs without distracting logos or labels. The new setting allows them to use Gemini's generation capabilities while still being compliant with best practices for AI transparency, thanks to the invisible metadata.

The timing of the rollout is also notable. As AI-generated content becomes increasingly common across social media, advertising, and entertainment, debates about labeling and transparency are more relevant than ever. Governments and regulatory bodies around the world are exploring rules for AI content disclosure. Google's decision to offer a visible watermark toggle, while keeping invisible markers, could position the company for future regulatory requirements.

Users who are part of the rollout can expect to see the Media Watermark option appear in their settings gradually. Google has not specified a precise schedule, but the company says the update will be complete within a few days. Once available, users can change their preference as often as they like, and the setting will persist across sessions.

It is also worth noting that the toggle does not apply retroactively to content created before the update. Existing generations with visible watermarks will keep them unless the user manually removes the watermark through editing tools. Google has not announced any feature that would automatically strip watermarks from previously created files.

For those with paid AI Ultra subscriptions in restricted countries, the setting provides a workaround to the regional limitation. However, the requirement to pay for the feature has drawn some criticism. Google has not explained why India, South Korea, and Vietnam are singled out, but it may be related to local laws or existing agreements with AI content providers.

The company has also updated its documentation to reflect the new setting, including guidelines for developers and enterprise customers. Work and school administrators will need to manage watermark policies through their Google Workspace settings, but they cannot enable the toggle for end users. This suggests that organizations may have different responsibilities when it comes to AI content labeling.

As Gemini continues to evolve, features like the Media Watermark toggle show that Google is listening to user feedback. The company has been iterating rapidly on its AI suite, adding new tools and improving existing ones. Nano Banana, Omni, and Lyria represent some of the most advanced generation capabilities in the consumer market, and the ability to control visible watermarks adds a layer of customization that was previously missing.

Looking ahead, it is likely that other AI platforms will follow suit or adjust their own watermarking policies. The balance between transparency and usability remains a central challenge for the industry. Google's approach — offering a user-controlled visible mark while preserving invisible traceability — could become a standard model. For now, users who have been frustrated by the stamped-on watermark have a reason to check their Gemini settings in the coming days.

Source:Digital Trends News