Google Opens SynthID Detector to Global Users
Google has made its SynthID Detector available worldwide in English. The web tool lets signed-in users upload images, video, or audio to check for hidden AI watermarks. It can identify content from multiple AI platforms, not just Google's.
Google's tool is a browser-based service that requires an account before use. After logging in, people can submit still images, sound recordings, or video clips for scanning. The system looks for concealed markers that supported AI systems add during generation, rather than relying on visible labels.
The detector is not limited to Google-made content. It also recognizes output linked to OpenAI, Nvidia, and Kakao. Apple plans to add support later this year, which would widen the range of platforms covered by the same watermarking system.
The worldwide release could make AI-content checks more accessible to journalists, fact-checkers, platform moderators, and ordinary users. If supported services keep embedding watermarks, the detector may help people trace synthetic images, audio, or video before sharing them. Its practical impact may remain uneven, however, because detection depends on compatible tools, intact files, and user sign-in. Broader adoption could gradually shape how audiences judge authenticity online, while also raising expectations that AI makers disclose and preserve such markers.