Google DeepMind's New SynthID Tool Watermarks AI Images With Imperceptible Pixels
- Google DeepMind released SynthID, a tool that watermarks AI-generated images with a digital imprint embedded in the pixels.
- The watermark is invisible to humans but can be detected by AI to identify synthetic content.
- SynthID is currently limited to Google's Imagen but may expand to other models to combat deepfakes.
- The tool marks images on a scale of high to low confidence based on the robustness of the watermark.
- While not foolproof, SynthID aims to address concerns over authenticity and copyright of AI-generated media.
TechCrunch 
Engadget 
Mashable 
The Messenger 
Gizmodo 
MIT Technology Review 
Washington Post 
The Verge 
