Google DeepMind's Backstory Speeds Up Media Fact-Checking
Backstory collapses what previously required five separate tools and logins into one workflow, cutting the most time-consuming part of image verification for fact-checkers.
Reporting from 1 source: GIGAZINE.
Google DeepMind's AI tool Backstory, announced in 2025, is accelerating fact-checking work at news organizations. The tool checks SynthID watermarks and C2PA metadata to detect AI-generated images, traces an image's online history, and compiles a cited report. It is currently available only through the Trusted Testers Program, with thousands of journalists and researchers testing it. India Today's six-member fact-checking team uses it for the first stage of verification.
Backstory is an AI tool developed by Google DeepMind. When a user uploads an image and enters basic information or questions, it automatically runs checks to determine whether the image was AI-generated. As part of its AI detection functions, Backstory checks SynthID, the watermark added by Google's image generation AI Nano Banana, as well as C2PA metadata, which embeds content credentials into images.
An AI agent compiles information on whether the image shows signs of tampering, whether it has appeared online before, and, if so, in what context it was used, generating a report with citations. As of the time of writing, it is only available through the Trusted Testers Program, and thousands of journalists, OSINT specialists, and information literacy researchers are reportedly testing it.
Synthesized by Yomimono from the 1 cited source below, including Japanese-language reporting where cited, then editorially reviewed before publishing.