Google has introduced a measure aimed at curbing the dissemination of false information.

On Tuesday, the company revealed an imperceptible but enduring watermark that will be applied to images to indicate their computer-generated origin.

This technological advancement, dubbed SynthID, integrates the watermark seamlessly into images produced by Imagen, a recent addition to Google’s suite of text-to-image generators, CNN reported.

Remarkably, the AI-driven watermark remains intact even if the image undergoes modifications such as the application of filters or adjustments to its colors.

The functionality of the SynthID tool extends beyond watermarking. It has the capability to analyze incoming images and determine the likelihood of their creation by Imagen.

This assessment is made through a three-tiered certainty scale: images are classified as either detected, not detected, or possibly detected based on the presence of the watermark.

Written by staff