Google Makes Visible AI Watermarks Optional, Keeps SynthID
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google announced Friday that users can remove visible watermarks from its AI-generated images, videos, and songs, while invisible SynthID watermarks and C2PA standard metadata remain in place for identification.
- Josh Woodward, Google's VP for Gemini, said on X the toggle applies to the Nano Banana, Omni, and Lyria models, and is live in Gemini and the Flow video editor with Search support coming soon.
- The on/off setting lives at Settings > Media Watermark, rolling out over the coming days once the feature is available.
- Google is also open-sourcing a new library called Credentio so developers can embed a local validation mechanism for AI-generated content in their own apps.
- Woodward framed the shift as balancing creative control with safety, noting Gemini and Search can still flag AI-generated images via the invisible layer.
- The move follows Anthropic's recent, heavily debated decision to watermark Claude-generated text and files to comply with EU regulations.
Why it matters: Creative and professional users gain flexibility to use Google's AI outputs without the visible mark, while the company preserves its transparency layer through invisible SynthID watermarks and C2PA metadata; the open-sourced Credentio library separately hands third-party developers a local way to verify AI content.
Ask SkimNews
