Google Opens SynthID to All, 1 Million Verifications Per Day

Google has made available a site for everyone to check whether an image, video, or audio file originates from a generative model. The company reports recording 1 million checks per day and has already linked this verification to its Gemini application and Chrome. This initiative is part of a landscape where OpenAI, Nvidia, and Kakao also support SynthID, while other players maintain their own standards and the tools remain improvable.

A claimed massive usage and a fragmented ecosystem

The SynthID verification is integrated into the Gemini application and the Chrome browser. Google states that 1 million daily requests are made to check content. OpenAI, Nvidia, and Kakao also support SynthID, and OpenAI offers its own verification site. Apple is expected to add support soon. Microsoft and Meta have their own standards for content labeling and verification. These tools are not infallible and often fail to identify content created by their developers' models.

A public site that verifies images, videos, and audio

The site launched by Google allows anyone to check whether an image, video, or audio file has been generated by AI. It is now accessible to all and supports numerous formats. For images, it accepts JPG, JPEG, PNG, BMP, WEBP, AVIF, HEIC, HEIF, TIFF, TIF, and GIF files. Videos can be submitted in MP4, MOV, and WEBM formats. For audio, the supported formats are WAV, MP3, OGG, FLAC, AAC, and M4A.

SynthID labels content from Google’s models and tools

The operation of the site relies on SynthID, introduced in 2023, which allows for determining whether a media file has been generated by AI. Google’s models like Nano Banana, Veo, and Lyria, as well as generation tools such as Gemini, Flow, ProducerAI, and Vids, apply this labeling during content creation. Before the public launch, the tool had been tested by journalists, media professionals, and researchers at Google I/O last year.