Google to Implement AI-Generated Image Labeling in Search Results

Google Search to start labelling AI-generated images

Google has announced that it will be rolling out a new feature in search that helps users distinguish between AI-generated images and images taken by humans. The move comes after the tech giant has joined the Coalition for Content Provenance and Authenticity (C2PA), a group of major brands that have developed the latest Content Credentials standard to prevent misinformation being spread online. Google said that the metadata for images will now include information like the origin of the image and where and how it was created. These labels can now be seen across Google apps like Search, Images and Lens. Once users click on the 'About this image' feature, they can see if an image was created or edited by AI. Besides this, the company is also working on bringing in AI content labelling into its ad platforms so that they comply with the C2PA policies when featuring AI-generated content. Google brings AI answers in Search to India, other countries YouTube is also reportedly looking at ways to identify and label videos created or edited by AI and is expected to announce news around this later during the year. There are however AI platforms that have declined to follow the C2PA standard like Black Forest Labs, the AI image and video generating startup behind Flux, the model that X's Grok uses. Google is also working on a watermarking tool called SynthID which was created by Google DeepMind to help identify AI-generated text, images, audio and videos online. Published - September 19, 2024 11:47 am IST Read Comments

Google is preparing a system to label all AI-generated content - Softonic

English The company will integrate the C2PA standard into many of its services and systems, including advertising Google has announced the implementation of a content authentication technology in its products. Over the coming months, the company will integrate the Coalition for Content Provenance and Authenticity (C2PA) standard into its search, advertising, and potentially YouTube services. The new measure aims to help users differentiate between images created by humans and those generated by artificial intelligence. Developed by a consortium of tech companies since 2019, the C2PA technology aims to combat the growing amount of AI-generated content, allowing the tracking of content origins and its editing history through metadata. At Google, it will be integrated into features like "About this image" in search, Circle to Search and Lens, showing information about whether an image has been edited with AI. Laurie Richardson, Vice President of Trust and Safety at Google, acknowledged the challenges of establishing content provenance in such a vast digital environment. "There is no magic solution for all online content," she stated, while emphasizing the importance of collaborating with other industry players to create sustainable solutions. With this initiative, Google aligns with other AI transparency measures, such as SynthID, a watermarking technology created by Google DeepMind. The C2PA standard will also be implemented in Google's advertising systems as part of its transparency policies. In the long term, it could extend to platforms like YouTube, especially for content captured by webcam. However, the adoption of this technology is not without problems. Many devices and editing tools are still not compatible with C2PA, making it difficult to maintain metadata. Additionally, AI image generators would need to adapt to include this information, which could exclude open-source models.