Google introduced a new feature for its Gemini app, allowing users to detect if a video was generated or edited by the company's AI tools. This update expands a similar capability previously available for images, aiming to increase transparency around AI-generated content. The feature utilizes Google’s proprietary SynthID digital watermarking technology, which embeds an imperceptible watermark directly into AI-created content.
To use the function, users upload a video file to the Gemini app and query if Google’s AI generated it. The application scans both the video and audio tracks for the SynthID watermark. It then provides a detailed response, indicating which specific segments contain AI-generated elements.
Current technical specifications limit video uploads to a file size of 100 MB and a maximum duration of 90 seconds. A critical limitation is that the tool only identifies content created by Google’s own AI models. The video verification feature is now available in all languages and countries supporting the Gemini app.