Google has updated its AI tools, allowing users to toggle off visible watermarks on their creations. However, invisible markers remain. The change could complicate the art of identifying AI-generated content.

The update includes a new 'Media watermark' setting in Gemini and Google’s AI video generator, Flow, which lets you turn off the ‘sparkle’ watermark that appears in the bottom-right corner of content generated with Nano Banana and Omni models. Invisible SynthID and C2PA watermarks will still be embedded, enabling verification through Gemini or Search.

While this update might make it harder to spot AI-generated content at a glance, many users have already been editing out visible watermarks. Other AI generators like OpenAI and Meta’s also don’t apply visible watermarks, relying on SynthID and C2PA for verification instead. Anthropic has recently announced applying invisible watermarks to both AI-generated text and images.

Google plans to roll out this new watermark setting to Search as well, though the toggle won't be available in countries that require visible watermarks.