Climate+Tech Launches Open Source AI Fact-Checker to Tackle Climate Misinformation
In an ambitious move to combat climate misinformation, Climate+Tech has unveiled an open source AI fact-checking system designed to assess climate-related claims against scientific evidence. The platform, known as Climate+Tech FactChecker, merges multiple AI models in a unique “advocate-mediator” architecture and has achieved 85% alignment with expert consensus in early testing.
“Misinformation is one of the greatest barriers to effective climate action,” said a spokesperson for Climate+Tech. “Our goal is to provide transparent, scalable tools grounded in science that can help people and organizations distinguish fact from fiction.”
The FactChecker system uses AI “advocates” to analyze claims independently, drawing on sources like the IPCC and peer-reviewed studies, while a “mediator” AI consolidates the findings into a final verdict. It supports both open-source and commercial models, allowing flexible deployment and customization for different users.
Initial tests on 170 expert-reviewed claims from Climate Feedback showed strong performance: over 80% agreement with experts on straightforward claims, and around 70% accuracy on more complex issues. The system identifies common misinformation tactics such as cherry-picking data and misrepresenting statistics.
Key features include transparent sourcing with citation links, rapid scalability, and the ability to process thousands of claims daily. Future plans focus on real-time source integration, multi-language support, and tools for community collaboration.
Climate+Tech invites scientists, developers, and fact-checkers to contribute via GitHub. “We’re building a community-driven ecosystem for trusted climate information,” the team said. “Open source is central to that mission.”
The project is licensed under AGPL for research and academic use, with commercial options available.
