Tool now available worldwide in English; OpenAI, NVIDIA and Kakao tools supported, Apple to follow
Google has made its SynthID Detector available to everyone, allowing anyone to check whether an image, video or audio clip was created with AI. The company announced the global rollout on Oct 7, with the tool available in English from day one.
The detector, at synthid.com, looks for SynthID's invisible watermarks. It covers content generated by Google's own AI tools and by partners including OpenAI, NVIDIA and Kakao. Google says Apple will be added soon.
In a blog post, Pushmeet Kohli, VP of Science and Strategic Initiatives at Google DeepMind, said that as AI tools improve and become easier to use, people need clear context about what they see online to decide what to trust.
SynthID launched in 2023. Google says it has since watermarked more than 180 billion images and videos, along with 240,000 years' worth of audio. An early version of the detector was released last year, but it was aimed at media professionals. Today's release removes that restriction.
The standalone tool adds to verification features Google already offers in Search, the Gemini app and Chrome. Together, those handle more than 1 million requests a day, the company said.
The announcement does not say how the detector treats content from AI tools that don't use SynthID. A result showing no watermark therefore wouldn't prove that a piece of media was made by a person.