I tested Google's new SynthID detector on AI slop. There's a big catch

PCWorld tested Google's SynthID detector, a tool designed to identify invisible AI watermarks in images, video, and audio content. The tool only detects watermarks from specific partners like Nvidia, OpenAI, and Kakao, meaning a negative result does not confirm media is human-made. SynthID can flag certain AI-generated content but falls short as a reliable authenticity checker, leaving a significant gap in detecting broader AI slop. Aw, have you seen that viral video of the little puppy nuzzling with a baby? It's absolutely adorable -- so adorable that it's probably AI, although not even Reddit seems to know for sure. Now Google has unveiled a site featuring a tool designed to spot AI-generated images, videos, and audio files, using a technology called SynthID to pinpoint invisible watermarks embedded in AI-generated content. Surely the SynthID tool can tell us whether that suspicious puppy video was created by AI, right? I plopped the ultra-cute video into the tool, waited a few seconds, and got the result: no AI watermarks detected. So, the video must be real, right?