Google Now Lets You Ask Gemini if an Image Was Created Using AI: How to Use

Google on Thursday announced the launch of a new feature in the Gemini app that allows users to verify whether an image was generated or edited using the company’s artificial intelligence (AI) tools. The Mountain View-based tech giant said the move is aimed at increasing content transparency using its digital watermarking technology, SynthID. This feature will also expand to support other formats beyond images, such as video and audio clips.

Synthetic ID

According to Google, SynthID is embedded in all images generated by its tools. It now provides verification functionality in the Gemini app to check if an image was generated by Google AI. The company says the visible watermark is available for free and professional users, while Ultra subscribers and enterprise tools will have the option to remove the visible watermark for professional work.

Google is also testing a verification portal called SynthID Detector with journalists and media professionals.

How to verify if an image was generated or edited using Gemini

1. Open the Gemini app and upload the image you want to inspect. Click the File icon to upload from your device or the Gallery icon to upload from your photo library
2. Once uploaded, enter a prompt such as “Was this created with Google AI?” or “Was this generated by AI?” in the text field.
3. Gemini will then check the SynthID watermark and return the response using its own reasoning
4. AI tools tell you what percentage of images were created using AI

According to Google, SynthID verification works on images generated using its proprietary AI tools and does not work on non-Google AI products. Notably, the company recently expanded this functionality to academia, researchers, and multiple media publishers.

Google says it will expand SynthID verification to support more formats in the future. Currently limited to images only, this feature will be extended to video and audio clips as well. Additionally, the company plans to add SynthID verification to more interfaces, such as search.

Notably, SynthID was first launched by Google DeepMind in August 2023 as a test project aimed at correctly labeling AI-generated content.