Google is set to release a new tool that helps determine whether an image was taken with a camera, edited with software like Photoshop, or created by a generative AI model. In the coming months, the company will roll out an updated “About this image” feature in its search results, enabling users to understand if an image has been altered or generated using AI technology. We’ll keep you updated on the rollout progress.
How Google’s Image Authentication Works
This new technology relies on a metadata standard promoted by the Coalition for Content Provenance and Authenticity (C2PA). C2PA aims to address the growing concern of AI-generated images by providing information on the image’s provenance. The authentication process involves creating a digital footprint that hardware and software systems can recognize. The initiative has been endorsed by major companies, including Amazon, Microsoft, Adobe, Arm, OpenAI, Intel, Truepic, and Google, although adoption has been slow. The integration of this standard into Google search results represents the first large-scale test of its effectiveness.
Expanding C2PA and Future Integration
Google helped develop the latest version of the C2PA standard, known as C2PA 2.1. The upcoming search feature will work in conjunction with a trusted list, allowing users to confirm the accuracy of metadata related to an image. For example, if metadata suggests that an image was taken with a particular camera model, the trusted list will verify that information. Google also plans to integrate C2PA into its advertising systems and is exploring its use on platforms like YouTube, adds NIX Solutions.
C2PA is currently supported by cameras from Leica and Sony, with Nikon and Canon pledging to add support soon. Apple and Google are expected to follow suit with their mobile operating systems. While Adobe Photoshop and Lightroom already support C2PA metadata, other software like Affinity Photo and Gimp do not yet offer this feature. Many online platforms are also not yet compatible with C2PA, but Google’s implementation may encourage wider adoption in the near future.