Google to Introduce C2PA Content Credentials to Help Users Identify AI-Generated Images
Google announced on Tuesday that there were new ways for users to find out whether an image was created using generative artificial intelligence (AI). While the tech giant has been working on developing internal tools to watermark AI-generated content, it also joined the Coalition for Content Provenance and Authenticity (C2PA) as a steering committee member in February. After working with other industry players to develop a new technical standard, the company is now integrating them into images accessible via its tools.
Google to Help Users Identify AI-Generated Images
The risk with AI-generated images is that many digitally created and enhanced images cannot be distinguished from real ones. Due to this, the problem of deepfakes has surged massively, where a realistic image generated by AI representing a person, place, or event is claimed to be real to spread misinformation.
Google stated in a blog post that in the first half of the year, it worked with other members of the coalition to develop a new version (2.1) of the technical standard dubbed Content Credentials. It is said to be more secure against different kinds of tampering and has a stricter technical requirement. This standard is now being added to images that can be looked up using Google tools.
The tech giant said Content Credentials will be integrated into images that show up on Google Images, Lens, and Circle to Search. This means when users to go the “About this image” section of an image, they can check the C2PA metadata confirming whether the image was created or edited using AI tools.
Additionally, Google said it intends to integrate C2PA metadata using its ad systems. This data will guide the company’s key policies and enforcement strategies in the future. Further, it is also looking at ways to relay C2PA information to viewers on YouTube to help them know whether the video was captured using a camera or was made digitally.
Notably, the tech giant has been developing an in-house watermarking technology for AI content called SynthID. This system was created by Google DeepMind and embeds the information within the pixels of the image in a way that it remains invisible to the eyes but shows up when checked using special tools.
