Connect with us

Google

AI Detective: Uncovering Google’s Gemini Video Creations

Published

3 months ago

on

Google’s Gemini app can check videos to see if they were made with Google AI

Google Enhances Gemini’s AI Verification Feature for Videos

Google has expanded Gemini’s artificial intelligence (AI) verification feature to include videos created or edited using Google’s own AI models. Users now have the ability to ask Gemini to determine if an uploaded video is AI-generated simply by asking, “Was this generated using Google AI?”

SynthID Watermark Detection

Gemini will analyze the visuals and audio of the video for Google’s proprietary watermark known as SynthID. Google states that the response provided will be more detailed than a simple yes or no. Gemini will specifically identify the times when the watermark appears in the video or audio. This capability was initially introduced for images and is now extended to videos that are created or edited with Google AI.

Challenges with Watermark Removal

While some watermarks can be easily removed, Google asserts that its own watermark is “imperceptible.” The ease of removing this watermark remains unclear, as well as how effectively other platforms will detect the SynthID information to label the content as AI-generated. Google’s Nano Banano AI image generation model within Gemini embeds C2PA metadata. However, the lack of consistent tagging of AI-generated content across social media platforms poses a challenge in detecting deepfakes.

Verification Process Details

Gemini’s video verification feature can process videos up to 100 MB and 90 seconds in length. This functionality is available in all languages and locations where the Gemini app operates, ensuring widespread accessibility.

See also  Google Hits Milestone with Over $400 Billion in Annual Revenue
Related Topics:
Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending