Google will now allow users to remove visible watermark from its AI generations

In Brief

Google announced connected Friday that it volition present let users to region a disposable watermark from its AI generations, including images, videos, and songs. The institution specified that this won’t impact the invisible SynthID watermark and C2PA standard-related metadata.

Josh Woodward, the company’s VP for Gemini, said successful a station connected X that this toggle volition beryllium disposable for Nano Banana, Omni, and Lyria models. He specified that the mounting to crook the disposable watermark disconnected volition beryllium disposable successful Gemini and Google’s video editor, Flow, with enactment for Search coming soon.

The alteration reflects an evolving attack to labeling AI-generated media, arsenic disposable watermarks tin often marque contented little utile for nonrecreational and originative work, but the request to place AI-generated contented remains.

“We’re striking a equilibrium present betwixt originative power and safety: portion the disposable watermarks are present optional, invisible SynthID watermarks and C2PA metadata are inactive being utilized for transparency. So you tin inactive usage Gemini oregon Search to spot if an representation was AI-generated,” helium wrote.

The diagnostic is rolling retired successful the coming days, and erstwhile available, users volition beryllium capable to spell to Settings > Media Watermark to crook the disposable people connected oregon off.

Google is besides open-sourcing a caller room called Credentio to fto developers embed a section validation mechanics successful their apps.

This comes aft Anthropic’s hotly debated move to see a watermark successful substance and files generated by Claude to comply with EU regulations.

Subscribe for the industry’s biggest tech news