Google has made its SynthID Detector publicly available worldwide in English, giving users a way to check images, video and audio for compatible digital watermarks associated with AI-generated material. The tool, available from October 7 at synthid.com, is designed to identify content made with Google AI tools and participating partners.
SynthID embeds imperceptible information into AI-generated files rather than placing a visible logo on an image or an overlay on a video. When the detector finds a compatible marker, it can indicate that the material has an artificial origin.
The result has an important limitation: SynthID Detector searches for supported watermarks; it does not determine whether every unmarked file was made without AI. A file that produces no match therefore cannot be treated as proof of authenticity. Its coverage depends on watermarking having been applied when the content was created.
Google says it has applied SynthID markers to more than 180 billion images and videos since the technology launched in 2023. The company also says its watermarked audio output adds up to 240,000 years of listening time.
The public website expands verification features already available in Google Search, the Gemini app and Chrome. Google says those services collectively process more than one million verification requests each day. An earlier version of the detector had been offered to a group of media professionals.
Supported content currently includes material produced with AI tools from the following companies:
- OpenAI
- NVIDIA
- Kakao
Google says support for content created with Apple AI tools is planned. Expanding participation to more AI providers would determine how much of the wider stream of synthetic media SynthID Detector can recognize.
