Google Photos Introduces AI Labels to Identify AI-Edited Images
Posted: Fri Oct 25, 2024 6:17 pm
Google Photos recently announced a new labeling feature to highlight images edited with artificial intelligence (AI) tools. This feature, designed to increase transparency, will provide AI-specific labels directly within the image metadata, making it easier to distinguish AI-enhanced photos from original ones. For now, these labels will be included for images edited within Google Photos using its proprietary AI tools, such as Magic Editor and Magic Eraser, but it's unclear whether images edited using third-party AI tools will also be flagged.
How AI Labeling Works in Google Photos
Whenever a photo is edited using Google’s AI tools, the app will embed information about the edits in the image metadata. This label is embedded in a way that keeps it intact even after cropping or blurring. However, metadata won’t carry over if the image is saved as a screenshot, since screenshots create new EXIF data. This labeling will follow technical standards set by the International Press Telecommunications Council (IPTC) rather than the C2PA standard, which is adopted by Meta and OpenAI.
In addition to metadata, Google Photos will also display AI labels within the app, making it easy to see if an image has been AI-edited directly in the photo’s information panel, labeled as “AI Info.” This section will identify the AI tool used and specify if the edit was generative or made through other editing tools.
Expanded Labels for Pixel-Exclusive Features
Google Photos will also label non-generative AI edits made with Pixel-exclusive tools like "Best Take" and "Add Me," providing an extra layer of information about sophisticated yet non-AI-based edits. This added transparency is part of Google’s push to make it easier for users to recognize digital alterations, building trust and accountability in an era of AI-generated media.
Through these changes, Google Photos is establishing a clear standard for identifying AI-generated and enhanced images, adding to users' control over content authenticity.
How AI Labeling Works in Google Photos
Whenever a photo is edited using Google’s AI tools, the app will embed information about the edits in the image metadata. This label is embedded in a way that keeps it intact even after cropping or blurring. However, metadata won’t carry over if the image is saved as a screenshot, since screenshots create new EXIF data. This labeling will follow technical standards set by the International Press Telecommunications Council (IPTC) rather than the C2PA standard, which is adopted by Meta and OpenAI.
In addition to metadata, Google Photos will also display AI labels within the app, making it easy to see if an image has been AI-edited directly in the photo’s information panel, labeled as “AI Info.” This section will identify the AI tool used and specify if the edit was generative or made through other editing tools.
Expanded Labels for Pixel-Exclusive Features
Google Photos will also label non-generative AI edits made with Pixel-exclusive tools like "Best Take" and "Add Me," providing an extra layer of information about sophisticated yet non-AI-based edits. This added transparency is part of Google’s push to make it easier for users to recognize digital alterations, building trust and accountability in an era of AI-generated media.
Through these changes, Google Photos is establishing a clear standard for identifying AI-generated and enhanced images, adding to users' control over content authenticity.
