Google will now allow users to remove visible watermark from its AI generations
Google will let users remove the visible watermark from AI-generated images, videos and songs made with certain Google models. The change aims to give creators more control over material used professionally or creatively, while retaining mechanisms intended to identify AI-generated media.
The optional setting will apply to Nano Banana, Omni and Lyria outputs in Gemini and Google’s Flow video editor, with Search support planned later. Invisible SynthID watermarks and C2PA metadata will remain, and the feature is due to roll out in coming days through Settings > Media Watermark; Google is also open-sourcing a validation library called Credentio.
- Google makes visible AI watermarks optional.
- Invisible provenance markers will remain.
- The setting arrives in Gemini and Flow.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Allowing users to remove a visible watermark gives people greater control over images they have generated and may make those images more usable in legitimate creative, commercial and accessibility contexts. Supporters can argue that an always-on mark can be intrusive or aesthetically damaging, while invisible identification benchmarks retain a means for platforms and investigators to assess whether a file was AI-generated. This approach seeks to balance practical usability with traceability.
The case against
Critics can argue that visible watermarks provide an immediate, public signal that an image was generated by AI, helping ordinary viewers assess material without specialist tools. Removing that signal may make deceptive sharing easier, particularly when images circulate outside platforms able to inspect invisible benchmarks. They may therefore prioritise clear, durable disclosure over user convenience, on the grounds that provenance should be understandable to the public at the point of viewing.