Google makes Gemini's visible AI watermark optional

Google announced it will let Gemini users turn off the visible watermark on AI-generated images, videos and songs, while leaving its invisible SynthID markers and C2PA metadata in place. The change covers content created in Gemini and Flow, Google's video editor, with support for Search coming later. Josh Woodward said on X that the toggle will apply to the Nano Banana, Omni and Lyria models. Once the feature arrives, users will find it under Settings > Media Watermark, and Google said it is rolling out in the coming days. The option will not appear in countries where visible AI watermarks are required. Woodward described the update as a balance between creative control and safety. Even if the visible watermark is switched off, users can still use Gemini or Search to check whether an image was AI-generated, because the invisible markers remain embedded. The move sits within a broader argument about how AI content should be labelled. Earlier reporting by Ars Technica said Google uses both SynthID and C2PA, with SynthID designed to survive transformations such as filters and cropping, while C2PA metadata is easier to strip by editing or screenshots. In a separate release, Google also said it is open sourcing Credentio, a C++ library for C2PA Content Credentials. The company said the same code powered nearly 40 conformant C2PA-enabled Google products and helped scale to tens of billions of generated assets across images, videos, audio files and documents, while local validation avoids sending media to cloud servers and reduces privacy, latency, bandwidth and file-size limits.