Google’s Gemini Now Verifies AI-Generated Videos with SynthID
Google
Gemini
AI Verification
SynthID
Watermark
Deepfakes
AI Detection
7
Watermark Wars
Media Hype
6/10
Real Impact
7/10
What is the Viqus Verdict?
We evaluate each news story based on its real impact versus its media hype to offer a clear and objective perspective.
AI Analysis:
While the SynthID watermark offers a solution, its vulnerability and the lack of widespread adoption of AI content tagging suggest a continuous arms race, limiting the overall long-term impact.
Article Summary
Google has expanded the capabilities of its Gemini AI assistant by introducing a video verification feature. Users can now utilize Gemini to check if a video they've uploaded was created with Google's AI models. The process involves asking the chatbot, “Was this generated using Google AI?”, which then scans the video for Google's proprietary SynthID watermark. Unlike simple 'yes' or 'no' responses, Gemini will pinpoint the exact timestamps within the video where the watermark appears, providing granular insight into the video's origin. This follows a previous rollout for Google AI-generated images. The feature’s efficacy, however, hinges on the difficulty of removing the SynthID watermark and the broader industry’s adoption of coordinated AI content tagging. This development comes amid growing concerns about deepfakes and the authenticity of digital media.Key Points
- Google’s Gemini can now verify video authenticity by scanning for the SynthID watermark.
- Users are prompted with the question, ‘Was this generated using Google AI?’ to initiate the verification process.
- The verification system identifies the specific timestamps where the SynthID watermark appears within the video.