Home / Technology / Google Lets You Erase AI Watermarks
Google Lets You Erase AI Watermarks
14 Aug
Summary
- Visible AI watermarks can now be optionally removed by users.
- Invisible watermarks and metadata still remain for verification.
- Other major AI generators also use invisible watermarking methods.
Google has introduced a new feature allowing users to remove visible watermarks from content created by its AI tools. A 'Media watermark' setting in Gemini and the company's AI video generator, Flow, enables users to toggle off the visible 'sparkle' watermark. This update aims to provide users with more control over the appearance of AI-generated media.
Despite the removal of visible watermarks, Google will continue to embed invisible watermarks, such as SynthID, and C2PA metadata. These hidden markers will still allow for the verification of AI-generated content through tools like Gemini and Google Search. This approach mirrors practices by other leading AI developers, including OpenAI and Meta, which also utilize invisible watermarking or metadata for content authentication.
This development comes as many users have already been editing out visible watermarks from AI-generated images. The option to disable visible watermarks is expected to roll out to Google Search as well, though it will not be available in regions mandating visible watermarks. This move signifies a growing trend towards less intrusive methods of signaling AI-generated content.
