Google has updated its Gemini and Google Flow platforms to allow users to toggle visible watermarks on and off for AI-generated images, videos and music tracks.
Josh Woodward, vice president at Google, announced that users can disable visible watermarking on images generated by AI Nano Banana, videos created with Omni and audio tracks generated via Lyria. However, visible watermarks will remain mandatory in regions with strict disclosure laws, such as China.
While Google did not disclose a detailed rationale for the policy change, Woodward characterized the update as a minor tweak. Industry observers suggest the decision stems from user frustration over watermark overlays obscuring simple AI-generated content.
The change brings Gemini in line with competitors like ChatGPT, which allows watermark-free AI content generation. However, both platforms continue to embed back-end metadata into media files to identify them as AI creations.
Users can modify their settings by navigating to Settings, selecting the Media Watermark option, and toggling the feature on or off. The preference will automatically apply to all future content generated on the platform.
Despite permitting users to hide visible watermarks, Google maintains a multi-layered security protocol in the background to preserve content provenance and prevent abuse.
All generated media will continue to feature SynthID, an invisible watermarking technology embedded directly into image pixels, audio spectra and video frames. This digital marker remains detectable by verification systems even if the file is edited, cropped or compressed.
Additionally, Google incorporates C2PA metadata, an industry standard established by the Coalition for Content Provenance and Authenticity, which logs the specific AI model used and the timestamp of creation within the file structure.
Source: The Verge, Tech Crunch, Techradar, Platform X
