Navigating the modern media landscape seems to be getting more difficult at an increasing pace. We’re saturated with synthetic content, inundated by outrage, and guided by algorithmic amplification. All this makes critical reading both more difficult and more necessary. This is the backdrop in which Fallacy Finder was born.
Rather than treating language as neutral information, Fallacy Finder approaches it as rhetoric: persuasion, framing, omission, emotional leverage, and argument. This is a tool designed to help users examine the logic of what they read more closely by identifying and highlighting possible logical fallacies in online articles and pasted text. The goal with Fallacy Finder is not to replace human judgment with a machine’s judgment. Instead, the web application supports real users by encouraging closer reading and fostering more reflective engagement with online content.
Reading With Suspicion, Not Cynicism
Logical fallacies are not merely technical mistakes in reasoning. In everyday life they often function as social tools: ways of redirecting attention, strengthening group identity, muddying evidence, or making weak claims sound stronger than they are. They appear in political speech, journalism, commentary, advertising, online debate, and ordinary conversation.
Making these patterns more visible helps users slow down, inspect claims, and ask better questions about how an argument is being built. With Fallacy Finder the hope is to get users thinking about what is being assumed, what is being smuggled in emotionally, and precisely where persuasion starts to become a substitute for proof.
How It Works
Users can submit a URL or paste text directly into the application. The system extracts the content, analyzes it with Large Language Models (LLMs), and returns a marked-up result that highlights passages associated with particular fallacies. These are then grouped and labeled so the user can review the structure of the argument in a more organized way.
The application is designed to be practical and exploratory at the same time. It can be used to examine news articles, opinion pieces, speeches, essays, or social media text. It is equally useful for teaching, research, media literacy, and personal experimentation.
AI as a Critical Companion
A great deal of AI discourse today is very binary. The two poles tend to consist of topics related to productivity, automation and acceleration vs uncovering exploitation, climate effects and existential dangers. While we clearly see all of these aspects of AI, we want to show a third way with these technologies. Specifically, with Fallacy Finder the idea isn’t to use AI technology to generate more content, but instead to help interrogate the content already surrounding us (preferably with local AI, running on our own hardware).
The project is less interested in artificial intelligence as a source of answers than as a tool for structured doubt. We want to ask whether machine analysis can be redirected toward critical literacy and help people read more carefully, recognize manipulation more clearly, and, hopefully, argue more honestly.