AI Creator Watermarks Go Public!

Google's SynthID Detector, a tool designed to identify AI-generated content, is now available to all users. This move expands access beyond journalists and researchers who previously used a waitlist-only version. The detector works by watermarking content created by Google's AI systems and those of its partners, including OpenAI and Nvidia. Apple is expected to integrate with the system soon, allowing its AI-generated media to be flagged.

The SynthID watermark is embedded upon content creation and is designed to remain intact through various modifications like cropping or filter application. Google reports that over 180 billion images and videos, along with 240,000 hours of audio, have been watermarked using the system. This initiative mirrors efforts by other tech giants like Microsoft and Meta, which also implement watermarking and detection mechanisms for their AI outputs.