Google Sharing SynthID Watermark for AI Text

Google is open sourcing its watermark technology for generative AI text outputs.

Google DeepMind shared its SnythID Text tech at GitHub, published a research paper on it and added documentation on it for developers, according to the company today.

SynthID Text is designed to allow developers to embed digital watermarks in text generated by AI without affecting quality, accuracy and creative output.

The watermark is imperceptible to humans but detectible for identification.

The tag is intended to address concerns that generative AI text could "contribute to misinformation and misattribution problems."

SynthID Text is not set up to "directly stop motivated adversaries from causing harm."

"However, it can make it harder to use AI-generated content for malicious purposes, and it can be combined with other approaches to give better coverage across content types and platforms," Google says.