Google on October 7 launched SynthID.com, a public website that lets anyone check whether an image, video, or audio clip was generated by AI1,2. The detector supports watermarks from all of Google's partners, not just its own Gemini models, closing a gap in the previous tool.
SynthID embeds an invisible signal in the pixels of AI-generated images and videos and in the waveform of AI audio. Google says Gemini alone has applied SynthID to more than 180 billion images and videos and 240,000 years of audio. ChatGPT-generated images also carry SynthID watermarks, but Google's prior detector would not have flagged them; the updated tool now does.
Before the site went live, the only way for users outside Google's trusted-tester program to check for SynthID was to query Gemini directly, a process Ars Technica described as clunky. SynthID.com accepts dozens of common file formats3.
ANALYSIS Opening detection to all comers, and extending it to partner watermarks, removes a bottleneck that limited SynthID's practical value as a provenance standard.