Launching Carboncredits.fyi: An AI-Powered Living Review of Carbon Credit Quality

Carbon crediting mechanisms are considered an essential tool for mitigating climate change, enabling governments and companies worldwide to offset part of their emissions. As carbon credit markets have grown rapidly, their effectiveness depends on four fundamental principles: additionality (ensuring projects wouldn't happen without carbon credit revenue), robust quantification (accurately measuring emission reductions), permanence (ensuring reductions aren't reversed), and avoiding double counting (ensuring each reduction is only claimed once).

A carbon credit (representing one ton of CO2) must meet these criteria to have real climate impact. The project must be additional - meaning it wouldn't have happened without carbon credit revenue. The emission reductions must be accurately quantified using robust methodologies. These reductions must be permanent rather than temporary. And finally, the credit must only be counted once toward climate goals. Without meeting these criteria, the system risks undermining both market integrity and climate action.

This is why continuous assessment of these credits is essential. As carbon markets evolve, so does the need for rigorous analysis. The launch of Carboncredits.fyi marks the publication of our global synthesis analysis, examining the actual climate impact of carbon credits across all these dimensions. We systematically assessed studies covering 2346 carbon mitigation projects to evaluate how well projects meet these fundamental criteria, focusing on additionality and robust quantification. Our research covers nearly one billion tons of CO2e - about one-fifth of all carbon credits issued to date. Our findings are clear: there are significant, widespread quality issues in carbon crediting.

With Carboncredits.fyi we're harnessing artificial intelligence to create a living, evolving assessment of carbon credit quality. Our interactive data explorer already allows you to filter observed reductions by project type, sector and geography. But we're going further: we're deploying AI to continuously scan and screen thousands of new academic papers, helping us rapidly identify and incorporate relevant new research. In the future, AI will help us automatically extract key statistics and findings from these studies, enabling near real-time updates to our systematic review.

As our AI capabilities expand, we'll add new features for visualizing and analyzing the data, making it easier for stakeholders to understand how different types of carbon credits perform against fundamental quality requirements. Join us as we build a more transparent, evidence-based carbon crediting system powered by the latest advances in artificial intelligence.