recently introduced Gemini 3, enhancing its AI chatbot with greater capabilities, such as generative UI. However, it isn’t this newfound ability that stands out as the most thrilling alteration to the Gemini model at this moment. In fact, one particularly intriguing advancement in Google’s AI lies in its newfound ability to determine if an image was produced by AI.
Granted, the concept of using AI to identify AI-generated content may seem somewhat absurd on its own. Nonetheless, possessing additional tools to ascertain whether something was crafted by AI is becoming increasingly vital as a growing number of AI solutions emerge, enabling users to generate convincing content that can easily be mistaken for something human-made. The capability to inspect images for an invisible watermark in AI-generated material is now being introduced for Gemini.
[embedded content]
This feature is being deployed as an advancement of Google’s earlier released SynthID Detector, which the company launched earlier this year. SynthID serves as a digital watermark incorporated into images, videos, and even audio and text that are AI-generated. This watermark has been integrated into several systems, including Gemini itself, and Google’s detection system can readily identify whether these concealed watermarks exist in the images that you upload
