About the project
Evidence.ninja is a next-generation tool that analyzes climate-related claims, identifies myths, and provides fact-checking using reliable sources. It uses a specialized AI model trained to recognize fallacies in the climate debate.
Ninja is based on peer-reviewed research (the CARDS framework—Computer Assisted Recognition of Denial and Skepticism) with transparent results and open methodology. The model has been thoroughly tested using a comprehensive taxonomy of arguments denying the severity of climate change, and all methods are open-source and verifiable. Evidence Ninja analyzes the full spectrum of arguments using a taxonomy of 103 categories. It covers everything—from legitimate scientific arguments to blatant disinformation. The model distinguishes between different types of claims rather than labeling everything as “disinformation.”
Ninja is capable of processing hundreds of thousands of messages, enabling it to provide an overview of the information landscape in near real time. Its results are processed by an analytical team that will regularly publish summary reports on the spread of misinformation in the climate debate.
The goal is to provide the public with a tool that allows them to verify climate-related claims, assess their credibility, and, where necessary, supplement them with targeted fact-checking based on reliable sources.
Evidence.ninja is the result of a collaboration between the Association for International Affairs (AMO) and Discourselab.ai from the United Kingdom. Discourselab.ai is a company founded by academics who have long been developing machine learning tools for the large-scale and rapid identification of myths, misconceptions, and misinformation in the climate debate. We have joined forces to create a project that aims to make advanced academic research accessible to the public.
