Google has opened its SynthID Detector for checking AI-generated content to the public, so anyone can now check whether an image, video or audio clip was made with AI. The tool went live globally in English this afternoon, which means it’s available in Ireland from today.
Until now, the detector was limited to journalists, researchers and other media professionals, who could access an early version through a waitlist after Google first showed it off in 2025. The public version recognises content made with Google’s own AI tools and with tools from its SynthID partners. Google lists those as OpenAI, NVIDIA and Kakao, and says Apple is joining soon.
What is SynthID?
SynthID is Google DeepMind’s invisible watermarking system. It doesn’t add a visible logo to a picture or rely on metadata, which is easily lost when a file is re-saved or screenshotted. Instead, it embeds a signal directly into the content at the moment it’s created. You can’t see or hear it, but the detector can pick it up.
Google says it has watermarked more than 180 billion images and videos since SynthID launched in 2023, along with 240,000 years of audio. Those are Google’s own figures, but they give a sense of how much watermarked content is already out there.
How to check if an image, video or audio clip is AI-generated
The process is simple enough. You sign in to synthid.com with a Google, Apple or OpenAI account, upload the file you’re unsure about, and the detector scans it for a SynthID watermark. Google hasn’t mentioned any cost attached to using it.
It sits alongside the verification features Google has already built into Search, the Gemini app and Chrome. The company says those now handle more than a million requests a day between them.
What SynthID Detector can’t tell you
This is the part to understand before you start lobbing every dodgy-looking image into it. SynthID Detector only finds SynthID watermarks. If something was made with an AI tool that doesn’t use SynthID, such as an open-source image model running on someone’s own computer, there’s simply nothing for it to find.
A negative result doesn’t mean a human made it. It only means the content doesn’t carry a watermark from Google or one of its partners. A positive result tells you far more than a negative one, and that distinction is going to trip people up.
That said, the partner list is the most significant part of this announcement. OpenAI already adds SynthID watermarks to images made in ChatGPT, and if Apple follows through, a large share of the AI content the average person comes across will be checkable in one place. That’s a far cry from when SynthID was effectively a Google-only system.
That being said, a fragmented approach is arguably worse leaving the onus on the public to know that not all AI images are detectable as easily as others.
Availability in Ireland
SynthID Detector is available now in Ireland and worldwide, in English only for the time being. Google hasn’t said when other languages will follow. You can read more in Google’s announcement of the public SynthID Detector.
