Google opens the source code of its SynthID text watermarking technology, which makes it easier to identify generated content. The code is available through the Google Responsible Generative AI Toolkit.
Today, we’re open-sourcing our SynthID text watermarking tool through an updated Responsible Generative AI Toolkit.
— Google DeepMind (@GoogleDeepMind) October 23, 2024
Available freely to developers and businesses, it will help them identify their AI-generated content. 🔍
Find out more → https://t.co/n2aYoeJXqn pic.twitter.com/4uRKYaz57Y
SynthID helps make AI-generated output discoverable by adding an invisible watermark to images, audio, video, and text as they are created.
”By open-sourcing the code, more people will be able to use the tool to watermark and determine whether text outputs have come from their own LLMs - making it easier to build AI responsibly," said Pushmeet Kohli, vice president of research at Google DeepMind.
Google claims that SynthID is already integrated into the Gemini chatbot. The feature does not degrade the quality, accuracy, creativity, or speed of the generated text. The tool can work with texts of three sentences or more, as well as with text that has been truncated, paraphrased, or modified. But it has difficulty working with short texts, rewritten or translated content, and even answers to factual questions.
“While SynthID isn’t a silver bullet for identifying AI-generated content, it is an important building block for developing more reliable AI identification tools.”