Google's SynthID detector is now open to everyone

Google DeepMind has opened the SynthID Detector, a tool for recognising AI-generated media, to everyone. It was previously a limited version available only to media professionals and researchers.

What does SynthID do?

SynthID embeds a watermark, imperceptible to the eye or ear, in images, video and audio produced by AI models. Google says that since launching the method in 2023 it has watermarked more than 180 billion images and videos, plus 240,000 years of audio. The detector checks whether a file carries that watermark.

According to TechCrunch, the site accepts image formats such as JPG, PNG and WEBP, videos in MP4, MOV and WEBM, and audio files such as WAV, MP3 and FLAC.

Which output does it recognise?

The detector only recognises content marked by Google and its partners: besides Google's tools such as Gemini, Veo and Lyria, OpenAI, NVIDIA and Kakao support SynthID, and Apple is expected to follow. TechCrunch notes that Microsoft and Meta use their own watermarking standards, and that such tools are not infallible, often failing to identify even content made by their makers' own models.

Verification features also exist in Google Search, the Gemini app and Chrome; Google says these channels receive more than 1 million verification requests a day.

What should brands and marketing teams do?

This goes beyond the facts in the sources: now that anyone can check content, AI-generated visuals are easier to spot. That does not mean using them is forbidden, but you should be ready for customers or journalists to ask questions.

• Keep a record of which tool produced and edited the images, video and audio used in campaigns.
• Check the labelling rules of ad platforms and publishers for AI-generated content.
• Add a production-method disclosure to contracts with agencies and freelancers.
• Remember the tool reads only supported watermarks; do not treat a 'not found' result as proof on its own.