Google lets users remove visible AI watermarks from Gemini content

Google announced on Friday that users will soon be able to remove visible watermarks from AI-generated images, videos, and music created with Gemini. The toggle, called "Media watermark," will be available in Gemini and the video editor Flow, with support for Search coming soon. The change applies to models Nano Banana, Omni, and Lyria, and is rolling out "except in countries where it's required by law to keep them," such as the EU and South Korea Source: Engadget.

What stays

Despite the removal of visible markers, Google emphasizes that invisible SynthID watermarks and C2PA metadata will remain embedded in content. Users can still verify AI generation by asking Gemini or Search Source: The Verge. Josh Woodward, VP of Gemini, stated, "We're striking a balance here between creative control and safety" Source: TechCrunch.

Industry context

This move aligns with broader industry trends. OpenAI and Meta do not apply visible watermarks, relying instead on SynthID and C2PA, while Anthropic recently added invisible watermarks to comply with EU regulations Source: The Verge. Google is also open-sourcing a library called Credentio to help developers embed local validation Source: TechCrunch.

Concerns

Critics argue that visible watermarks are a key tool for quickly identifying AI content. Gizmodo notes that C2PA metadata is easily stripped, and SynthID, while robust, is not designed to withstand adversarial attacks, with commercial services already claiming to bypass it Source: Gizmodo. The decision is seen as another step toward eroding the public's ability to recognize AI-generated media at a glance.