Google to let users remove visible watermarks from AI content

Google is preparing to let users turn off visible watermarks on AI-generated images, videos and other content, giving creators greater control over how their work is presented.

The option will initially be introduced for content generated by Google’s Nano Banana, Omni and Lyria models through a simple toggle. Similar controls may later come to Gemini and the AI video platform Flow.

Removing the visible watermark will not disable AI identification. Google said its invisible SynthID watermark and C2PA metadata will remain, allowing AI-generated content to be verified through appropriate tools.

The move is aimed at balancing creative freedom and AI transparency, although critics warn that removing visible markers could make AI-generated content harder for ordinary users to identify. The feature will also be restricted in countries where visible AI labelling is legally required, including parts of the European Union.

Google’s decision comes days after Anthropic announced machine-readable watermarks for Claude-generated content, highlighting the contrasting approaches major AI companies are taking toward content transparency.