NoticeDomain for sale and contact us at hello@emerging.news

Google's upgraded SynthID detects AI generated media with invisible watermarks from multiple major platforms.

Google has upgraded its SynthID tool with the new SynthID Detector. This enhancement expands its capability to identify AI generated media beyond Google's own models. The detector can now scan images, videos, and audio files originating from OpenAI, NVIDIA, and Kakao, with plans to include content from Apple soon. SynthID helps users identify AI generated or manipulated media by detecting invisible watermarks embedded by Google's AI models. Since its 2023 launch, Google watermarked over 180 billion images and videos, plus 240,000 years of audio content.

To use the SynthID Detector, users visit synthid.com and log in with a Google, Apple, or ChatGPT account. After uploading a media file, the tool indicates if a SynthID watermark is detected and if the content is designated AI generated. It is crucial to note that the SynthID Detector functions by identifying these specific watermarks, not by broadly detecting all AI generated content. If AI generated media lacks a SynthID watermark, the tool cannot determine its AI origin.