Google's SynthID watermark is hard to break, but it doesn't solve AI misinformation

Google's SynthID watermarking technology makes it difficult to remove identification markers from AI-generated images, according to the company's research. The watermark persists through common image modifications including compression, cropping, and color adjustments. However, the technology does not address how to verify the authenticity of non-watermarked images or prevent the spread of AI-generated content that lacks the watermark.

Next story