Google has announced a new update for the Gemini app that allows users to edit images directly within conversations, whether the images are AI-generated or uploaded from a phone or computer.
The company stated that the update began rolling out gradually today and will expand to users in over 45 languages worldwide in the coming weeks.
This development builds upon the experimental image editing model in the AI Studio platform, which Google introduced last March.
On its blog, the company highlighted being the first to place AI image editing capabilities directly in users' hands.
Through the new tool, it's now possible to perform a series of multiple edits in sequential steps, allowing for easy background changes, replacement of elements within images, and addition of new details.
Google notes that this interactive approach provides users with more accurate and contextual responses by blending text and images within the conversation.
At the same time, however, this feature has sparked significant debate due to its ability to remove watermarks from images.
In this context, the company explained that the tool could be used, for example, to try different hair colors on a personal photo or to generate a children's story with accompanying images starting from a simple idea.
How to Use the New Image Editing Feature in Gemini
To take advantage of the new editing tools within Gemini once available in your region, you can follow these steps:
1. Update the Gemini app to the latest available version.
2. Open a new chat window within the app.
3. Select an image from your device or use a previously generated image.
4. Use the editing tools such as:
- Adjusting lighting and contrast.
- Cropping or resizing the image.
- Adding filters or text.
- Saving the final image or sharing it directly within the conversation.
Although these capabilities open the door to creative uses, they also raise questions about the risks of deepfakes.
Therefore, Google has affirmed that all images edited using Gemini will carry an invisible watermark, and it is currently testing the addition of visible watermarks as well to enhance transparency.
