Google has unveiled SynthID, a new watermarking tool designed to identify content generated by artificial intelligence. The tool embeds imperceptible digital watermarks into AI-generated images, audio, text, and video, helping users distinguish between AI-created and human-created content.

SynthID works by embedding watermarks directly into content at the moment of creation. For images and video, the watermark remains intact through common modifications like cropping, filtering, and compression. In audio content generated through Google's Lyria music model or Notebook LM's podcast feature, the watermark is inaudible yet resistant to noise and compression. For text, SynthID adjusts the probability scores of generated tokens to create a watermark that doesn't affect output quality.

Users can verify content through Gemini by uploading files to check for SynthID watermarks. Google has also launched the SynthID Detector portal for journalists and media professionals to verify content authenticity. The tool is being integrated across Google's generative AI consumer products to promote transparency and trust in AI-generated content.