Google is allowing users to remove visible watermarks from AI-generated images, videos and songs while retaining invisible provenance markers.

Google is giving users more control over the visible watermarks placed on content generated by its artificial intelligence tools.

The company is introducing a new setting that allows users to remove the visible watermark from images created with its Nano Banana image generator. The watermark, which appears in the bottom-right corner of generated images, had been a point of criticism among some users.

The new option is being rolled out in Gemini and Flow, while Google plans to make the feature available in Google Search’s AI Mode soon.

The change will not be limited to AI-generated images. Google will also allow users to remove visible watermarks from videos created with Gemini Omni and songs generated using Lyria.

There is one exception. Google said visible watermarks will remain in countries where keeping them is required by law.

Despite making visible watermarks optional, Google is not removing the underlying systems used to identify AI-generated content.

The company will continue using SynthID, its invisible watermarking technology, alongside C2PA metadata. These technologies are designed to provide information about the origin of AI-generated content even when a visible watermark has been removed.

As a result, users will still be able to ask Gemini whether an image was generated using AI. If the system identifies the image as AI-generated, it can provide that information even without a visible watermark.

The move gives users greater flexibility over how AI-generated content appears while maintaining Google’s existing approach to content provenance and identification.

The update also highlights Google’s effort to balance the growing demand for cleaner AI-generated media with the need to make synthetic content identifiable.