Google is adding two new features to image search to reduce the spread of misinformation, especially now that artificial intelligence tools have made it easier to create photorealistic fakes.
Alphabet Inc.’s first new feature is called ‘About This Image’, which allows you to see what your image and similar images look like when they are first indexed by Google, where they first appear, and elsewhere online. Provides additional context, such as whether it was displayed. Its purpose is to help users pinpoint the original source of the information while also helping to contextualize the images with disproving evidence that may have been provided by news outlets.
Google marks as such all AI-generated images created with its own tools, and works with other platforms and services to ensure that the same markup is added to the files they publish. Midjourney and Shutterstock are among Google’s participating publishers, and their goal is to ensure that all AI content appearing in search results is flagged as such.
Image provenance online is becoming an increasingly big problem in the AI era, with several startups working on verification and authentication tools. For example, Microsoft-backed Truepic Inc. offers a system that ensures images have not been manipulated from capture to delivery. Google’s new features rolling out this year are relatively low-tech, but with good industry support, they can have a much more positive impact.
