Google is introducing new features to help identify AI images in Search and elsewhere. In this way, fake artificial intelligence images used for malicious purposes will be detected.
Google is taking the first critical step to combat the potential for misinformation from AI-generated images. Among the announcements made at the I/O 2023 developer conference was the detection and identification of images created by artificial intelligence. The firm is introducing two new features for this: “About this image” and more background information about the image itself, and direct identification with “artificial intelligence generated” labeling.
The company says the new features are an extension of its existing work to include more visual literacy and help people more quickly assess whether an image is reliable or artificially generated. Of course, these features brought by Google do not prevent images created by artificial intelligence for the purpose of misinformation, but at least they facilitate the detection of these images.
“About this image”
The first new add-on coming “in the coming months” will offer more information about images in the “About this image” feature. With this feature, the user will be shown where the searched image or similar appears for the first time online. Of course, where else the image was seen will also be shown. This second part may include such things as news sites or fact-checking sites; this may prompt web searchers to learn more about the image, including how it may have been used in misinformation campaigns.
According to Poynter research conducted in 2022, Google states that 62 percent of people now encounter false information on a daily or weekly basis. Google hopes to fix this issue with the “About this image” feature. Google is proactive by promoting this tool before the internet is flooded with subtle AI-generated images (for example, Pope Francis wearing a puffer jacket).
“Created by artificial intelligence”
Later this year, users will be able to access the feature by right-clicking or long-pressing an image in the mobile and desktop Google Chrome web browser. However, this feature will initially be limited to the US and will only be available in English.
Meanwhile, it introduces a second feature that makes it easier to detect AI-generated images. For now, images produced by Google’s artificial intelligence tools will be labeled “artificial intelligence created”. Google says many publishers will adopt the feature, including Midjourney, Shutterstock, and others. Google said this feature will also be available “in the coming months.”