By Abigail Mbah
Google has opened its SynthID Detector tool to the public, making it simpler for anyone to check whether images, videos or audio files carry invisible watermarks that mark them as AI-generated.
The company announced the wider release on 7 October 2026. The detector is available globally in English through a dedicated website at synthid.com.
Users can upload media and see if it contains SynthID watermarks applied by Google or several partners, including OpenAI, NVIDIA and Kakao. Support for Apple is expected soon.
SynthID, first introduced by Google DeepMind in 2023, embeds digital signals that are imperceptible to people but detectable by the tool.
Google says it has so far watermarked more than 180 billion images and videos, plus the equivalent of 240,000 years of audio content.
Until now the detector was mainly available to journalists, researchers and media professionals for testing.
The public version joins existing verification features already built into Google Search, the Gemini app and Chrome, which together handle more than one million checks each day.
Pushmeet Kohli, vice president of science and strategic initiatives at Google DeepMind, said the move aims to give people clearer context about the media they encounter online as generative AI tools become more widespread and realistic.
The tool only identifies content that carries a SynthID watermark. The absence of a watermark does not prove a file is authentic, since not every AI system uses the technology and some content can be altered or created without it.
The release comes as concerns grow about realistic deepfakes, cloned voices and other synthetic media circulating on social platforms.