Google is enhancing content transparency with its Gemini app by enabling users to verify videos created by its AI models. This feature allows for the identification of AI-generated content through a unique identifier known as SynthID, embedded using Google’s proprietary digital watermarking technology. Users can simply upload a video and inquire if it was produced with Google AI, receiving a detailed analysis—beyond a yes or no answer—that specifies which segments contain AI elements. However, the verification process is limited to files up to 100 MB and 90 seconds long, making it ideal for short clips rather than full-length films. While Google’s SynthID technology has marked over 20 billion AI-generated pieces since its inception, this tool is specifically for content generated within Google’s ecosystem. Therefore, it cannot detect AI-generated videos from non-Google models, emphasizing that it is not a universal AI detection tool.
Source link
Share
Read more