Anyone can now check if a picture or video was made with Google's AI
Google opened its SynthID Detector to everyone. It looks for a hidden mark that Google and some partners, including OpenAI, put in AI-made files.
Is that photo real? A free tool from Google can now help answer that question for some pictures and videos.
Google has opened its SynthID Detector to everyone. Before, only an early version existed, and it was made for media professionals. Google says the new tool is "available globally in English starting today." The company shared the news on October 7.
Here is how it works. Since 2023, Google has added a hidden mark to files made with its AI. You cannot see it or hear it. Google calls these "imperceptible watermarks". A watermark is a hidden sign that shows where a file came from.
Google says it has marked "over 180 billion images and videos". It has also marked 240,000 years of audio.
Now anyone can upload a picture, a video or a sound file. The detector checks if the hidden mark is there. It works for files made with AI from Google or its partners. Google names OpenAI, NVIDIA and Kakao. It says Apple is coming "soon".
The same kind of check is already built into Google Search, the Gemini app and Chrome. Google says those features handle "over 1 million requests daily."
Why does this matter? Fake pictures and videos spread fast online. A quick check can help you decide what to trust before you share something with friends or family.
For now, the tool works in English.
Open it and check the claim yourself. That is the point.