On Friday, Google revealed its decision to permit users to eliminate the visible watermark from AI-generated content, such as images, videos, and music. The company clarified that this change will not affect the hidden SynthID watermark and the metadata related to the C2PA standard.
Josh Woodward, Vice President for Gemini, communicated in a post on X that this option would be available for the Nano Banana, Omni, and Lyria models. He noted that users can disable the visible watermark through Gemini and Google’s video editor, Flow, with plans to add Search support shortly.
This shift represents a changing strategy in how AI-generated media is labeled. Visible watermarks can often detract from the utility of content for professional and creative endeavors, yet the importance of identifying AI-generated material persists.
“We are finding a balance between creative freedom and safety: while users can now choose to remove the visible watermarks, the invisible SynthID watermarks and C2PA metadata will continue to provide transparency. This means users can still verify if an image was created by AI using Gemini or Search,” he explained.
The rollout of this feature is set to occur in the upcoming days. Once it’s available, users can navigate to Settings > Media Watermark to toggle the visible watermark on or off.
Additionally, Google is releasing a new open-source library named Credentio, which allows developers to integrate a local validation system within their applications.
This development follows Anthropic’s controversial decision to add watermarks to text and files generated by Claude to satisfy EU regulations.
