The News, Explained
Google made SynthID Detector available globally in English on October 7, 2026. The tool looks for SynthID digital watermarks in image, video and audio files to help identify AI-generated media and its source. A digital watermark places a signal in the content that people are not meant to see or hear directly. Source
Its stated scope is media made with AI from Google or participating providers and marked with SynthID. Google names OpenAI, NVIDIA and Kakao as current partners and says Apple support is coming soon. The formats named in the announcement are images, video and audio; text is not listed. Source
A user can submit a file to the Detector and check for the watermark signal. Google says it has watermarked more than 180 billion images and videos and 240,000 years of audio since introducing SynthID in 2023. Those are Google’s operational figures. Source
The Detector is a separate web tool, while related verification features are also built into Google Search, the Gemini app and Chrome. Google says those features now handle more than one million requests a day. The confirmed launch condition is worldwide access in English; the announcement gives no schedule for other languages. Source
OYOPICK’s Take
OYOPICK thinks the result is best read as a search for a supported provenance signal. As more providers join, the range of content that can be checked under the same system can grow. The announcement does not provide grounds to treat it as a universal detector for every generator and format. This is our interpretation of the stated coverage.
A widely used provenance signal could reduce the burden of guessing from an image or video alone. The presence of a signal and the truth of the content remain different questions. We hope the Detector adds more providers and languages while making the meaning of each result easy to understand. That is a conditional editorial hope, not a confirmed roadmap.