Google Gemini Now Lets Users Remove Visible AI Watermarks

Google has introduced a new option that allows Gemini users to remove the visible watermarks from AI-generated images and videos. The change gives users more control over how generated content looks, while Google’s invisible AI identification systems remain in place.

Visible Watermarks Can Be Turned Off

Previously, images and videos generated with Gemini displayed a recognizable Gemini icon in the bottom-right corner, making it immediately apparent that the content had been created by AI.

Google now allows users to disable that visual marker through the Gemini settings. Under Media watermark, users can choose whether the visible watermark is turned on or off.

The option currently applies to Google’s Nano Banana image model, the Omni video model and Lyria for AI-generated music.

Invisible Watermarking Remains

Turning off the visible logo doesn’t make the content completely unmarked.

Google executive Josh Woodward said invisible SynthID watermarks and metadata will remain embedded in generated images. The system is intended to preserve a way of identifying AI-generated content even when the visible indicator has been removed.

The visible watermark option will also remain unavailable in countries where AI-generated content must legally carry a visible watermark.

A Growing Debate Over AI Provenance

Google’s timing is notable, as AI watermarking has become a major topic following Anthropic’s announcement that it plans to embed invisible watermarks in AI-generated text.

The broader debate centers on content provenance—giving people a reliable way to determine whether an image, video or piece of text was created or modified by AI.

Google’s approach attempts to balance the two sides: users receive greater creative control by removing the visible Gemini logo, while invisible identification through SynthID and metadata remains intact.