VLDBench: A large-scale benchmark for evaluating Vision-Language Models (VLMs) and Large Language Models (LLMs) on multimodal disinformation detection.

Your personal detective in the world of disinformation. Built with ❤️ during EUDIS Defence Hackathon.

Codes and Recipe for Compiling and Analyzing "Four Shades of Life Sciences" Data Set

Adversarial AI engine that reverse-engineers video content to detect narrative manipulation, emotional coercion, and logical fallacies. Powered by Google Gemini 3.

An open-source platform for sharing, verifying, and discussing global events with transparency, aiming to reduce misinformation through community-driven fact-checking.

These notebooks analyze daily trends in online news coverage, examining news volume, topic distribution, source reliability, disinformation tactics, check-worthy claims, and visual-text alignment.

Repository zur Masterarbeit Retrieval Augmented Disinformation Detection

Lensmen is an investigative analysis tool. You submit any claim, question, or piece of evidence, and Lensmen applies 16 structured analytical lenses to find nuttinesses — things in the record that don't add up, contradict each other, or sit oddly against what should be true. Every finding points somewhere specific. Nothing is left as a dead end.

Source Code for the Bachelor's Project: Hybrid Small Language Models for Accurate Multimodal Disinformation and Misinformation Analysis

Bharat AI is a next-generation misinformation detection framework combining local LLMs and transparent reasoning. It offers multilingual verification, context awareness, and a user-friendly interface for real-time fact-checking across India.

🛡 Detect cognitive exploitation in texts, URLs, and files using pattern matching and optional LLM analysis with Seithar Cognitive Defense.