Google has launched an image generation feature in Gemini 2.0 Flash, which allows the model to create and edit images on its own. Users quickly discovered that the model is able to generate and animate realistic images of famous people, create deepfakes, and remove watermarks from stock images, TechCrunch reports.
Gemini doesn't just remove watermarks, it also seamlessly fills in the gaps left behind by removing them. While there are other AI-powered tools that can do similar tasks, Gemini 2.0 Flash has the advantage of being free.
The image generation feature in Gemini 2.0 Flash is currently marked as "experimental" and "not for production use" and is only available in Google's developer tools, such as AI Studio. It has some drawbacks: the models struggle with semi-transparent watermarks and very large watermarks.
Some copyright holders will likely be unhappy with the lack of restrictions on Gemini 2.0 Flash usage. Incidentally, Gemini's most popular competitors, such as Anthropic's Claude 3.7 Sonnet and OpenAI's GPT-4o, explicitly refuse to remove watermarks. Claude even calls removing a watermark from an image "unethical and potentially illegal."