There was a time when spotting AI-generated content was almost embarrassingly easy. Six fingers, melted backgrounds, dead-looking eyes, and robotic voices usually gave the game away. Generative AI has moved past many of those mistakes, and our eyes and ears are having a much harder time keeping up.
To be fair, plenty of AI-generated content is harmless fun. A ridiculous clip of a cat dancing or some impossible meme is hardly a crisis. The problem starts when the same tools are used to fake disasters, impersonate people, or manufacture convincing audio and video. Google has a new tool that could help.
SynthID Detector is now available to everyone
Google is making its SynthID Detector available globally in English. Upload an image, video, or audio file, and the tool checks for the imperceptible SynthID watermark embedded into AI-generated media. Google says SynthID has already been added to more than 180 billion images and videos, plus 240,000 years of audio.
The bigger development is that the detector is no longer useful only for Google-made content. It can also recognize supported AI media from partners including OpenAI, Nvidia, and Kakao, with Apple joining soon. Google has already baked similar verification tools into Search, Chrome, and Gemini. Gemini, for example, can inspect videos for SynthID across both their visual and audio tracks.
We need tools like this now
Synthetic media is getting frighteningly convincing. AI voices can laugh, whisper, pause naturally, and carry emotion convincingly enough to fool listeners. SynthID Detector is not a universal truth machine. It can identify supported content carrying SynthID, but an unmarked file is not automatically authentic. Still, giving everyone an easy verification tool is a good move. We are rapidly reaching the point where “it looks real” simply isn’t good enough anymore.
RELATED COVERAGE: