Google Gemini Allows Users to Toggle Visible AI Image Watermarks

Google has officially introduced a new feature allowing users to remove visible watermarks from images generated by its Gemini platform. This update provides users with the flexibility to toggle these markings on or off through a new setting, moving away from the previous standard where such branding was automatically applied to all outputs.

According to the original publisher, the announcement was shared by Google Labs Vice President Josh Woodward via a post on X. The new functionality extends beyond just static images; it also applies to videos created via the Omni tool and music generated using the Lyria model. Furthermore, Google intends to expand this option to include the AI Mode feature currently integrated into Search.

While users can now choose to hide the visible watermark, Google confirmed that this change does not affect the underlying security measures embedded within the content. Invisible identifiers, specifically SynthID and C2PA metadata, will remain embedded in the files to ensure that the origin of the content can still be verified.

The company has clarified that this setting is not universal. In regions where local laws mandate that AI-generated content must contain visible disclosures, the watermarks will remain compulsory and cannot be disabled. This ensures compliance with international regulatory frameworks while offering more control to users in other jurisdictions.

For Malaysian users and the broader digital community, the ability to remove visible branding highlights the ongoing evolution of AI transparency. As generative media becomes increasingly sophisticated, reliance on visible logos has proven to be an inconsistent method for identifying synthetic content. While the removal of these marks offers greater creative freedom for content creators, it underscores the importance of the hidden metadata that maintains the integrity of digital media provenance in an era of rapid AI adoption.