Google Lets Users Hide Gemini's AI Watermarks
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google added a "Media watermark" setting in Gemini and its Flow AI video generator, letting users remove the visible "sparkle" watermark from content made with the Nano Banana and Omni models.
- Even with the toggle off, Google will continue embedding invisible SynthID watermarks and C2PA metadata, per Josh Woodward, VP of Google Labs, Gemini, and AI Studio — content can still be checked via Gemini or Search.
- Google plans to extend the toggle to Search but won't launch it in countries that mandate visible watermarks on AI-generated content.
- OpenAI and Meta already skip visible watermarks on their AI image generators — OpenAI relies on SynthID and C2PA, while Meta uses its own Content Seal standard.
- Anthropic announced this week it is applying invisible watermarks to both AI-generated text and images.
- Many users were already editing out visible watermarks before Google's policy change, according to the report.
Why it matters: Dropping the visible "sparkle" makes it harder to spot AI-generated images at a glance, pushing detection entirely to invisible SynthID and C2PA markers that require Gemini or Search to verify. With OpenAI, Meta, and now Google all skipping visible watermarks, the industry is consolidating around invisible provenance — but only in jurisdictions that don't require visible marks.
Ask SkimNews
