Google now lets users verify whether a video was created or edited with Google AI directly in the Gemini app.
The verification system is powered by SynthID, Google’s invisible watermarking technology that identifies AI-generated content.
Users can upload short videos, ask Gemini to detect AI involvement, and receive clear results indicating whether AI markers are present in the clip.
The rapid advancement of artificial intelligence in recent years has made AI-generated photos and edits spread across the internet. With increasing precision by AI tools, it is more challenging nowadays to differentiate between authentic videos and synthetic edits. Therefore, if you are looking for transparency in the evolving AI-based world, Google has a solution for you.
The tech giant has introduced a useful verification feature in the Gemini app. This feature helps users detect whether a video is created or modified using Google’s AI models. Gemini scans it for SynthID watermarks.
Whether you’re a journalist, content creator, educator, or simply a viewer, the verification tool enables you to analyze suspicious clips and prevent misinformation. If you’re unaware of how to use this feature and what the feature can really do, below are the steps that you can rely on.
Apparently, the verification process may seem complicated to players. However, Google has made the process extremely simple and user-friendly. Still, following the steps correctly is important. Here’s how you can verify AI-generated photos and videos in Gemini:
Begin the process by launching the Google Gemini app on your mobile. You can also use Gemini on the web. Sign in with your Google Account and verify the authentication as prompted.
Once you are on Gemini, tap on ‘Add file’ and upload the video or photo you want to verify. You can select clips from device storage, gallery, or Google Drive. However, there are a few criteria to maintain. The file must be under 100 MB, and less than 90 seconds (for videos).
If your selected clip qualifies all the criteria, ask Gemini to verify the video. For that, simply upload the clip and give a prompt like, “Is this video (photo) created by Google AI?” or “Check if the video (photo) is AI-generated.”
After you give Gemini the prompts, wait for a while. The AI needs a few seconds to detect the hidden SynthID watermark. If it is found, Gemini will inform you which part of the video has the watermark. If no watermark is found, Gemini will clearly state that the video or the photo has no Google AI SynthID.
Also Read: How to Use Gemini Live API Native Audio in Vertex AI: Step-by-Step Guide
The feature is powerful, undoubtedly, but users need to understand that it has boundaries. Gemini verifies videos created or edited with Google’s AI technology. If you upload a clip generated by any third-party AI tool without a SynthID, Gemini can’t confirm it as AI content.
So, below are the restrictions that users should keep in mind while using Gemini for detecting AI-video or photos:
Only short videos under 90 seconds can be detected.
Uploaded files must be below 100 MB.
Daily verification limits are there. You can’t try the feature as many times as you want.
Gemini can detect only videos generated by Google AI models. Videos or photos by other AI tools can’t be detected.
Indeed, Gemini significantly improves media transparency. Still, users shouldn’t consider it as the ultimate tool for judging AI videos.
Now you are probably wondering what if Gemini can’t detect the video or photo, but it is actually generated by AI tools. Well, that’s a critical question indeed. Sometimes, journalists and educators have to verify the content for authenticity to prevent misinformation from spreading. If Gemini can’t help you, you have to find an authentic source before circulating it on the internet or in the classroom for students.
Alternatively, if Gemini detects the video (photo), you can easily skip that as a source of authentic information. Therefore, Google's AI detection tool contributes to maintaining healthy information circulation on the internet.
Also Read: Wipro Goes All-In on Google Gemini AI With Global Rollout and Bengaluru Experience Zone
If you are still confused about why the Google verification system is important, it is more than just a convenience tool. The newest Gemini feature allows users to stay safe in the era of manipulation and misinformation. If Google AI tools generate a video, Gemini will accurately detect the AI-generated visuals.
With this feature, Google aims to maintain trust and transparency in the AI-driven world of media and technology. By allowing users to verify whether videos were created or edited using Google AI, the company is trying to build trust in online content. With just a few steps, users can confidently verify authenticity and stay informed in the evolving digital landscape.
1. What exactly does Gemini verify in a video?
Ans: Gemini checks for Google’s invisible watermark, called SynthID, on all videos. This shows whether a video has been produced or altered using Google’s AI services.
2. Can Gemini confirm if a video is fake?
Ans: Gemini does not verify whether a video is altered. Instead, it checks whether a video contains a Google SynthID watermark. Therefore, a video without a watermark could still be an AI-generated video produced by another method.
3. What are the file requirements for verification?
Ans: There are a few criteria to match. Your file should be less than 100 MB and less than 90 seconds long.
4. How do I start the verification?
Ans: Nothing complicated. Upload the video in Gemini and give a prompt, "Is this video produced by Google AI?" or apply the shortcut command @synthid.
5. Does Gemini indicate which parts of the video are AI-generated?
Ans: Yes, it does. If Gemini detects Synthetic ID, it marks the areas of the video that have the watermark.