Google says 180 billion images and videos now carry SynthID watermarks as detector goes public

Google has opened its SynthID Detector to the public, allowing anyone to check whether an image, video or audio file was created by Google's AI or that of its partners. The tool identifies invisible watermarks embedded in content from models including Gemini, Veo and Lyria, and accepts JPG, PNG, MP4 and MP3 files.

SynthID is not a universal AI detector. It only flags content that carries a SynthID watermark; other detection systems share this limitation, typically verifying only their own provider's watermarks. For text detection, OpenAI developed its own separate system.

Beyond Google, OpenAI, Nvidia and Kakao also use SynthID for media, and Apple is expected to join soon according to Google. SynthID detection is additionally built into Google Search, the Gemini app and Chrome, where Google says one million verification requests arrive daily. The company states that more than 180 billion watermarked images and videos exist to date.

Why it matters: Developers and product teams can now verify SynthID-tagged content directly through the public detector or through integrations already present in Search, Gemini and Chrome, though the tool cannot identify AI-generated media that lacks a SynthID watermark. Apple's expected adoption would extend the watermarking scheme to another major platform.