With AI images getting increasingly better, it’s getting harder to tell them apart from real life stuff. Google has an idea in mind, although we have yet to see whether it will work fine.

Google Photos is adding new features to clearly label images that have been edited using its AI tools, such as Magic Editor and Magic Eraser. This move aims to increase transparency and help users understand when AI has been used to alter an image—it comes very in handy now that generative image tools are widespread and Google itself has added a bunch of AI-powered editing tools to the Photos app itself.

Previously, these AI editing tools were only noted through metadata embedded in the image file. Now, users will see a clear indication within the Google Photos app itself, alongside other image information like file name and location. This labeling will apply to edits that involve generative AI, which creates new image content, as well as non-generative AI used in features like “Best Take.”

Google cites its AI Principles as a guiding force behind this decision. The company plans to continue exploring further solutions to enhance transparency around AI’s role in image editing.

Unfortunately, this isn’t necessarily a foolproof way of telling whether something was actually made by AI or not. This takes the image’s metadata, which should tell if an image was AI-produced or modified, and the metadata in an image is editable. This is something that’s even shown off as a disclaimer within the app. Furthermore, if it was taking visual cues or indicators in an image, then it’s as simple as cropping out or editing out those indicators to remove any sign of AI in the image—at least signs that would allow Google to automatically detect if an image is AI-modified or not. Still, it’s better than having nothing at all.

Source: Google