They use Gemini 2.0 Flash to erase water marks with copyright

Users of social networks and forums such as X and Reddit have detected that some people are using the artificial intelligence model (AI) Gemini 2.0 Flash to eliminate water brands from author rights or copyright in deponiable in image banks such as Getty Images.

At the end of last January, the company implemented its GEMINI 2.0 Flash models, with multimodal capabilities of text, programming, video and spatial understanding; and a higher performance, since it is able to double the speed of Gemini 1.5 Pro.

Users are using this model, which can generate and edit images of images natively, to create images of famous people with copyright without limitations, eliminating water marks from existing photographs and violating copyright.

This has been denounced by some people on platforms such as X and debate forums such as Reddit, where they have shared their experience with Gemini 2.0 Flash, which “is incredible to edit images with simple text indications “, As a person identified as Tanay Jaipuria has pointed out. He has also pointed out that it is possible to eliminate water marks from the images and place his in place.

Techcrunch, who has compiled the tests of these users and has advanced the concern that this can be for image banks such as Getty Images, has indicated that the function of generating images of Gemini 2.0 Flash is offered as ‘experimental’. This means that it still fails to erase semi -transparent water marks or those that cover a large part of the images. In addition, it is “not suitable for the production ‘of images and is only available in the tools for Google developers, such as Ai Studio.

This media has also recalled that other models of competing brands, such as Claude 3.7 Sonnet of Anthropic and GPT-4o of OpenAi explicitly refuse to eliminate water marks when considering this as something “not very ethical and potentially illegal” according to Claude.