Google has announced a significant change to the way it labels content created using its artificial intelligence tools. Users will soon be able to turn off the visible watermark that appears on AI-generated images, videos and music, giving creators greater control over how their finished content looks.
The change was announced by Josh Woodward, Google’s vice president for Gemini, who said the company is trying to balance creative freedom with the need for transparency. The visible watermark will become optional, but Google will continue using hidden identification technology to indicate that content was generated by AI.
Visible AI Watermark Can Be Disabled
Until now, Google’s AI-generated content commonly displayed a small sparkle-style mark that indicated its AI origin. With the new setting, users can choose whether the visible watermark appears on their creations.
The option is being introduced for content generated using Google's Nano Banana, Omni and Lyria models. The feature is rolling out through Gemini and Google’s Flow video editor, while Google plans to bring similar functionality to Search in the future.
For creators, the change could be particularly useful when AI-generated material is being used for professional projects, presentations, marketing campaigns, entertainment content or creative work where an additional logo may affect the visual appearance.
Invisible SynthID Will Remain
Removing the visible watermark does not mean that Google's AI identification system is being removed.
Google will continue embedding its invisible SynthID watermark into supported AI-generated content. The company also retains C2PA-related metadata, allowing the origin of AI-generated media to be checked even when the visible indicator is switched off.
This means a person viewing an image may not immediately see an AI label, but Google's technology can still help determine whether the content was generated by AI.
Google has previously described SynthID as a system designed to help identify AI-generated content and promote transparency. The company said in May that SynthID had already been used across a huge volume of generated images, videos and audio.
Why Google Is Making the Change
Visible AI labels can make it immediately clear that content was created using generative AI, but they can also be inconvenient for users who want a clean final image or video. Google’s latest move appears to acknowledge this concern while maintaining an invisible layer of provenance.
The decision could also increase competition among AI platforms over how much control users have over generated content. At the same time, it raises questions about whether removing visible labels could make it harder for ordinary users to distinguish AI-generated media from authentic material.
Google's approach attempts to address that concern by keeping SynthID and metadata-based identification active even when the visible watermark is disabled.
