In a bid to curb on spread of deepfakes and concerning AI-generated images, Google has officially integrated an image verification feature into its Gemini AI app. The core of this new feature is SynthID, which is Google’s proprietary digital watermarking technology. SynthID works by embedding invisible markers directly into the pixels of images created or modified by Google’s AI models.
These markers are designed to be imperceptible to the human eye, ensuring that the AI image appears normal during viewing, but are instantly readable by the Gemini system. This allows the AI to state whether an image originated from Google’s own generative tools or an actual camera source.
Google Gemini gets SynthID watermarking
“Starting today, we’re making it easier for everyone to verify if an image was generated with or edited by Google AI right in the Gemini app, using SynthID, our digital watermarking technology that embeds imperceptible signals into AI-generated content,” stated Google in its press release.
According to the company, this feature is a significant step toward transparency, building on the foundation laid since SynthID’s 2023 launch, which has already watermarked over 20 billion AI-generated images.
“SynthID adds an invisible digital watermark to an AI-generated image (or video segment). The watermark doesn’t change the image or video quality. It’s added the moment content is created, and designed to stand up to modifications like cropping, adding filters, changing frame rates, or lossy compression,” says Google. “The watermarks are embedded across Google’s generative AI consumer products, and are imperceptible to humans – but can be detected by SynthID’s technology,” it adds.
How to use this feature
Users can access this verification with simple conversational prompts. One can upload an image to the Gemini app and ask questions like, “Was this created by Google AI?” or “Is this AI-generated?”
Gemini scans the image for the hidden SynthID signal before providing an answer.
However, the tool can only identify content created or modified using Google’s specific AI models. If an image originates from a competitor’s AI model or an independent developer, Gemini cannot verify its source. Hence, if someone has to test an image generated by ChatGPT or Grok AI, Gemini won’t be able to identify whether that image is real or created by AI.
Going ahead, Google says that “images generated by Nano Banana Pro (Gemini 3 Pro Image) in the Gemini app, Vertex AI and Google Ads will have C2PA metadata embedded, providing further transparency into how these images were created.”
