Google Lets Gemini Users Hide AI Watermarks but Keeps a Digital Trail

Google Lets Gemini Users Hide AI Watermarks but Keeps a Digital Trail
Google is loosening one of the most visible signs that media was made by AI, betting that hidden provenance tools can carry the burden of disclosure without spoiling creative work.

The change, announced Friday, allows users of Google’s generative tools to remove visible watermarks from AI-made images, videos and songs. The option is rolling out for Nano Banana, Omni and Lyria through Gemini and Google’s Flow video editor; Search support is expected to follow.

The policy shift reflects a practical clash: prominent watermarks can limit the use of generated material in professional and creative settings, while regulators, platforms and audiences still need a reliable way to identify synthetic media. Google’s answer is to make the on-image label a user choice, rather than abandoning attribution altogether.

Josh Woodward, Google’s vice president for Gemini, framed the move as an attempt to preserve both aims: “We’re striking a balance here between creative control and safety.” Even when the visible mark is switched off, he said, invisible SynthID watermarks and C2PA metadata will remain embedded for transparency, allowing Gemini or Search to check whether an image was AI-generated.

Once the rollout reaches users, the setting will sit under Settings > Media Watermark. Google is also open-sourcing Credentio, a library intended to help developers build local validation into their own apps.

The decision lands amid a broader argument over how conspicuously AI output should be labelled. It follows Anthropic’s contested decision to watermark text and files generated by Claude in response to EU rules—an indication that the industry agrees on the need for provenance, even as it diverges sharply over whether people should see it at first glance.

Continue reading https://foxvector.com/stories/01a01e49-c11e-19b9-70c2-12a9de4bde0d