Google added a Media Watermark setting that lets Gemini and Flow users switch off the visible watermark stamped on their images, video clips, and music tracks. Josh Woodward, who leads Google Labs and the Gemini app, announced the option in a Friday post on X, saying it covers the Nano Banana image model, the Omni video model, and the Lyria music model. The setting is rolling out over the coming days and applies going forward to newly created content.
Turning the toggle off changes only what a viewer can see. The invisible SynthID mark and C2PA metadata continue to ride inside every file no matter how the toggle is set, and Google’s detection tools such as Lens, AI Mode, and the “About this image” feature keep reading that layer the same way.
Google also open sourced Credentio, a C++ library that checks C2PA credentials without calling Google’s servers, and said nearly 40 of Google’s own products already rely on it. The visible toggle will not appear in countries with legal disclosure mandates.
For anyone reviewing AI-generated media by eye, the signal that used to be visible at a glance now depends on running a verification tool instead.
Search Engine Journal’s Matt G. Southern reported the change on August 17, 2026, citing Josh Woodward’s post on X.
