Google DeepMind Open-Sources SynthID Text for Detecting AI Content via Responsible AI Toolkit

Oct 23, 2024, 04:33 PM

Google DeepMind has open-sourced its AI text watermarking tool, SynthID Text, making it freely available to developers and businesses through an updated Responsible Generative AI Toolkit. The tool, which allows for watermarking and detecting text generated by large language models, has been used with Google’s Gemini chatbot and is intended to help identify AI-generated content. The technology introduces additional information into the text, improving the ability to trace synthetic content without altering the user experience. The release coincides with the publication of a paper in Nature describing SynthID Text, encouraging collaboration on responsible AI development.