Google will now allow users to remove visible watermark from its AI generations
TechCrunch Ivan Mehta ● Covered by 2 sources
Google now lets users hide the visible watermark on its AI images, videos and songs. The hidden SynthID mark stays, so AI can still be checked.
Based on reporting by TechCrunch, Ivan Mehta — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Google is giving people a new switch: the visible watermark on AI-made images, videos and songs can now be turned off. The company said the change applies to its Gemini family, including Nano Banana, Omni and Lyria.
The important catch is that the watermark isn’t disappearing entirely. Google says the invisible SynthID watermark and C2PA-related metadata will still be there, which means the company still wants a way to tell AI content from human-made work. Josh Woodward, Gemini’s VP, framed it as a compromise between creative control and safety.
That compromise is now built into the product. Users will be able to find the setting in Gemini and in Flow, Google’s video editor, under Settings > Media Watermark. Support for Search is coming later, so the rollout isn’t finished yet.
Google is also opening up a new library called Credentio, which it says will let developers add local validation into their own apps. The move lands after Anthropic’s much-discussed decision to add a watermark to text and files generated by Claude to meet EU regulations. Google, for its part, seems to be saying that visible labels are optional, but proof of origin is still non-negotiable.
My take — AI-written commentary, not fact-checked reporting
This is the right call, because a watermark that ruins the work was always a weird way to “protect” it. Google is basically admitting that creators want usable outputs, not giant digital post-it notes, while regulators still get their precious trail of metadata. The industry keeps pretending transparency and utility are enemies; they’re only enemies when companies design badly.
Read more about this at: TechCrunch
Related stories
Google's SynthID watermark is hard to break, but it doesn't solve AI misinformation
Ars Technica · 1 month ago ·
41
Making it easier to understand how content was created and edited
Google DeepMind · 3 months ago ·
35