Google DeepMind expands content provenance tools with SynthID and C2PA

Google DeepMind is expanding its toolkit for transparency and content verification by integrating SynthID — its digital watermarking technology for AI-generated content, which has already been applied to over 100 billion images and videos as well as 60,000 hours of audio. The company is also expanding support for the C2PA Content Credentials standard, which shows how media was created and altered, including with AI. The Pixel 10 is the first smartphone with support for Content Credentials for images in the native camera app, and in the coming weeks the technology will be extended to video on the Pixel 8, 9, and 10. In the Gemini app, SynthID verification has been added for images, video, and audio — with over 50 million uses worldwide — and this capability will later appear in Search and Chrome. Additionally, C2PA Content Credentials checking is being added to Gemini, allowing users to see whether content is an unaltered original from the camera or has been edited. Google is launching a new API for AI content detection on the Google Cloud Gemini Enterprise Agent platform to identify content created by both Google and other popular models. The company continues its industry collaboration: OpenAI, Kakao, and ElevenLabs are integrating SynthID into their AI content, while Meta will label camera images using C2PA in Instagram.