A public-interest tool for detecting deepfakes and verifying media authenticity using multiple AI detectors and human verification.
TrueMedia is a public-interest technology developed at Georgetown University for exposing deepfakes, manipulated media, and false information shared online. It checks the authenticity (AI-generated or not) of image, audio, and video, and verifies the veracity (true or false) of text content. The tool aggregates analysis from multiple AI detectors and human verification, providing transparent reporting of accuracy metrics and uncertainty. It is non-partisan, open source, and aims to reduce barriers to trustworthy media analysis in an AI-saturated information environment.
Key Features
check_circleChecks authenticity of image, audio, and video
check_circleVerifies veracity of text content
check_circleAggregates analysis from multiple AI detectors
check_circleHuman verification included
check_circleTransparent reporting of accuracy metrics and uncertainty
check_circleOpen source infrastructure
check_circleNon-partisan and public interest focused
Use Cases
lightbulbJournalists verify the authenticity of images and videos before publishing, reducing the risk of spreading deepfakes.
lightbulbSocial media moderators assess flagged content for AI manipulation, ensuring platform integrity.
lightbulbResearchers analyze patterns of AI-generated misinformation to inform policy recommendations.
lightbulbFact-checkers use the tool to quickly determine if a viral claim is supported by real media or fabricated.
lightbulbElection officials monitor for deepfakes targeting candidates, protecting democratic processes.
lightbulbEducators teach media literacy by demonstrating how to detect manipulated content using open tools.
lightbulbCybersecurity professionals investigate online fraud involving fake audio or video evidence.
