Google allows Gemini users to decide for themselves whether images, videos, and music pieces created with artificial intelligence carry a visible Google logo as a label. Until now, the company automatically printed the marking on every file; it can now be turned off in the settings. However, those using Gemini through a Google Workspace business account still cannot remove the labeling.
Google hides the visible AI logo
Those who generate an image in Gemini with the Nano Banana model, edit a video with the Omni tool, or compose a song with Lyria have previously had a small, permanently visible mark automatically embedded in the file. This marking was intended to show at first glance that the content was created with artificial intelligence. Since August 14, 2026, this visible mark can be turned off via a new switch called “Watermark for Media.” Gemini head Josh Woodward announced the change via the microblogging service X and describes it as a balance between creative freedom and the traceability of AI content.
Importantly: The invisible technical marking SynthID and additional provenance data remain in every file, even if the visible logo is absent. Therefore, anyone checking a file via the Gemini app or in the future through Google Search will still recognize an AI origin. Google announced that it will also extend the switch to Google Search results in the coming weeks; a specific date for this has not yet been set.
What this means for office work
Those who occasionally create an illustration for a presentation, an internal newsletter, or a social media tile using Gemini at work could not remove the small AI mark without cropping the image afterward. The new switch now automatically takes care of this: an image generated with Gemini for the team meeting title graphic becomes a picture without a visible indication of AI origin, while the invisible marking remains in the background. Those generating such images directly in Google Docs can now even find a dedicated function for AI graphics and diagrams there.
For all those using Gemini through a business Google Workspace account, however, nothing changes: the visible logo remains displayed in any case, and the new switch cannot be activated in business and educational accounts. Those working with a private Google account should still check whether internal guidelines for labeling AI content in customer materials or official communication explicitly require a visible mark.
Availability, costs, and how to adjust the setting
The feature is part of the free Gemini usage and does not require an additional subscription. Exceptions are users in countries where visible markings are legally required, such as China; in India, South Korea, and Vietnam, the switch is only available with a paid AI Ultra subscription according to Google’s own help page. In Germany, Austria, and Switzerland, the setting can be used without restrictions.
Here’s how to find the switch:
- Open Gemini in the browser at gemini.google.com or the Gemini app.
- Tap the gear icon for settings in the bottom left.
- Select the menu item “Watermark for Media.”
- Set the slider to “Off.”
The setting currently applies to the image generator Nano Banana, the video editor Omni, and the music generator Lyria in the Gemini app, as well as in the separate video tool Flow.
It will be crucial whether companies allow their employees to use AI images without visible labeling in official communication. The EU transparency obligations for content clearly marked as AI, which have been in effect since August 2026, remain unaffected by Google’s switch and continue to apply regardless. How quickly visible markings can lose significance was recently demonstrated by the case of the AI image function in Google Earth, which Google disabled within a day after fake explosion images appeared.
