In response to the rise of AI-generated content, Google is updating its tools to enhance transparency regarding the origin and modification of images in search results.
The tech giant is collaborating with the Coalition for Content Provenance and Authenticity (C2PA) to develop a global standard aimed at identifying and certifying AI-generated content.
This partnership has led to the release of C2PA's content credentials version 2.1, designed to improve security and prevent metadata manipulation for images, videos, and audio files.
These credentials will allow users to determine whether a digital file was created by a camera, edited with software, or entirely generated by AI.
Google has integrated C2PA's content credentials into key products, including Google Search and soon Google Ads. Users can access detailed origin information by clicking on the three vertical dots above an image in search results.
If the content contains C2PA metadata, users will see information about the device or software used for capturing or modifying the image. For instance, if a photo was taken with a specific camera model, this will be verified through a 'Trusted List' to ensure accuracy.
According to Laurie Richardson, Google's Vice President of Trust and Safety, the use of these authenticity signals will expand to other products, influencing advertising policies to combat misinformation.
YouTube, also owned by Google, will follow TikTok's lead by implementing content credentials to automatically label AI-generated videos, providing users with clarity on the nature of the content they view.
This initiative is part of a broader strategy by Google to combat online misinformation, including the launch of SynthID in 2023, a digital watermarking tool developed by Google DeepMind that detects and tracks AI-generated content.