Google's SynthID Watermark Is Resilient but Not a Cure for AI Disinformation

The scale of AI-generated media is staggering: a Stanford and USC research collaboration estimates that it took humanity 149 years after the invention of the camera to create 1.5 billion images, yet generative AI achieved the same volume in just 18 months. Google recently announced that its tools have produced over 100 billion AI images and videos in only a few years, and it is expanding adoption of its SynthID watermarking technology through new partnerships. While SynthID is designed to be hard to remove, experts note that labeling AI content alone does not solve the broader challenge of disinformation.