⚡
Brief IA
›

Google Makes Watermarking Optional for Its AI Content

🤖 Models & LLM·Tom Levy·

Google Makes Watermarking Optional for Its AI Content

Google Makes Watermarking Optional for Its AI Content
⚡
Key Takeaways
1The setting to make the visible watermark optional will arrive in the coming days, via Settings > Media Watermark.
2The option will be available in Gemini and Flow, with support for search announced as next.
3Invisible watermarks SynthID and C2PA metadata remain applied, according to Google.
💡Why it matters — Google maintains SynthID and C2PA metadata and states that Gemini or search can verify the origin of images while making visible marking optional.
⚡Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisis ton rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

📄
Full Analysis

Google plans to make the visible watermark on content generated by its AI models optional, while retaining the invisible SynthID footprint and C2PA metadata. The rollout is expected in the coming days, with access specified in Gemini and Flow, and support for search anticipated. A library called Credentio has been opened to integrate local validation on the developer side. This development comes as Anthropic has chosen to add a watermark to the outputs of Claude to comply with European regulations.

Google launches Credentio and prepares for deployment in the coming days

Google is launching Credentio, a library designed for developers to integrate a local validation mechanism into their applications. The new watermarking option is set to be deployed in the coming days, with an adjustable setting available, once released, via Settings > Media Watermark. This development comes as Anthropic has decided to add a watermark to the texts and files generated by Claude to comply with EU requirements.

The option will be offered in Gemini and Flow, with search to follow

According to Josh Woodward, the option to disable the visible watermark will be accessible from Gemini and the video editor Flow, with support for search announced as forthcoming. He specifies that the setting will pertain to the Nano Banana, Omni, and Lyria models.

Optional visible watermark, SynthID and C2PA maintained

Google announces that it will allow the removal of the visible watermark from images, videos, and songs generated by its models, without altering the use of the invisible SynthID watermark or C2PA-compliant metadata. Josh Woodward discusses the search for a balance between creative control and security, making the visible watermark optional while retaining invisible identification mechanisms. He also notes that Gemini or search can be used to verify if an image was produced by AI. This change is presented as an adaptation of watermarking, as visible watermarks can make certain professional and creative uses less practical, while still maintaining the requirement for identification.

⚡

Brief IA — L'actualité IA en français

L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.