"Gemini will scan the video's visuals and audio for Google's proprietary watermark called SynthID. The response will be more than a yes or no, Google says. Gemini will point out specific times when the watermark appears in the video or audio. The company rolled out this capability for images in November, also limited to images made or edited with Google AI."
"Google calls its own watermark "imperceptible." Still, we don't yet know how easy it will be to remove, or how readily other platforms will detect the SynthID information and tag the content as AI-generated. Google's Nano Banano AI image generation model within the Gemini embeds C2PA metadata, but the general lack of coordinated tagging of AI-generated material across social media platforms allows deepfakes to go undetected."
Gemini's verification feature now accepts videos made or edited with Google's AI models and answers queries like, "Was this generated using Google AI?" The system scans visuals and audio for Google's proprietary SynthID watermark and highlights specific times when the watermark appears. The capability was previously applied to images and remains limited to content produced or edited with Google AI. SynthID is described as imperceptible, but watermark removal risks and cross-platform detection challenges remain uncertain. Google's Nano Banano model embeds C2PA metadata. Verification supports videos up to 100 MB and 90 seconds and is available in all Gemini app languages and locations.
Read at The Verge
Unable to calculate read time
Collection
[
|
...
]