You can now turn off Google Gemini's visible watermarks

Google has officially changed the rules for how AI-generated media looks on the web. According to reporting by The Verge, the tech giant is now allowing users to toggle off the visible watermarks that previously appeared on images, videos, and music created with its AI tools. This update applies to the new Media watermark setting found in both Gemini and Google's AI video generator, Flow.

For creators who have been frustrated by the sparkle icon in the bottom-right corner of their outputs, this is a significant quality of life improvement. The toggle specifically affects content generated using Google's Nano Banana and Omni models. When you switch the setting off, that persistent visual branding disappears, giving your generated assets a cleaner, more professional appearance right out of the box.

However, do not mistake this aesthetic change for a complete removal of provenance tracking. Josh Woodward, the vice president of Google Labs, Gemini, and AI Studio, clarified that the invisible SynthID watermarks and C2PA metadata remain embedded in the background. These digital fingerprints are designed to persist even after the visible sparkle is gone, ensuring that the origin of the content can still be verified by systems that know where to look.

This shift signals a maturation in how major tech companies approach AI content labeling. The industry is moving away from aggressive, visible branding toward a more sophisticated infrastructure of invisible verification. It suggests that Google believes the future of trust lies in technical metadata rather than visual stigmatization, allowing creators to use AI tools without permanently marking their work as synthetic to the naked eye.

The implications for professionals are complex. On one hand, removing visible watermarks makes AI-generated assets more versatile for commercial use where branding consistency is key. On the other hand, the persistence of SynthID means that the line between human and AI creation remains technically traceable. This creates a dual reality where content looks authentic but carries a hidden digital signature that can be detected by compliant platforms and tools.

For entrepreneurs and marketers, this update highlights the importance of understanding the tools you use. You can now produce cleaner assets for social media or client presentations, but you must remain aware of the underlying metadata. As platforms increasingly adopt C2PA standards, the invisible watermarks may become just as important as the visible ones for compliance and copyright management.

What this means for you: You should update your workflow to test the new watermark settings in Gemini and Flow. Use the toggle to see how your specific use case benefits from the cleaner look, but always verify the output metadata if you are working in regulated industries. Try this prompt to audit your own generated content: "Analyze the attached image for C2PA metadata and SynthID signatures, and explain what provenance information is embedded in the file headers."

The broader trend here is clear. AI generation is becoming indistinguishable from human creation in appearance, but the technical infrastructure for tracking it is becoming more robust. This balance between aesthetic freedom and digital accountability will likely define the next phase of content creation across the web.