AI
Gemini lets users hide visible AI watermarks
Gemini users can hide visible watermarks on AI images, videos, and music, but invisible SynthID and C2PA markers remain.
Google Gemini users can now hide the visible logo added to AI-generated images, videos, and music. The change was announced on August 14, 2026, by Google executive Josh Woodward and reported by TechRadar.
The setting applies to images created with the Nano Banana model, videos made with Omni, and songs generated by Lyria. It is not available in countries where visible AI watermarks are required by law, including China.
Woodward called the change a “papercut fixed,” suggesting the overlaid Gemini logos had become an annoyance for users who wanted to share or publish clean-looking creations. Visible watermarks can now be toggled in Gemini and Flow, while Google Search is expected to get the option next.
“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.”
— Josh Woodward, Google executive
The removal affects only the visible label. Google will continue embedding SynthID, an invisible watermarking system, along with C2PA metadata that identifies AI-generated content inside the files. Gemini can also analyze an uploaded image and identify it as AI-generated when those invisible signals are present.
How to disable Gemini’s visible watermark
On the web, click the cog icon in the lower-left corner of Gemini, then select Media watermark and switch the setting off. TechRadar says the control was not yet visible in Gemini’s mobile apps, although Google is expected to add it to their settings.
The change brings Gemini closer to ChatGPT, which does not add visible watermarks to images. It also creates a clearer split between presentation and provenance: users can remove the logo, but the generated file can still carry machine-readable signals that reveal its origin.