About

Frontier AI research and development is accelerating, with new capabilities emerging at a breakneck pace. These systems offer unprecedented opportunities for innovation and scientific progress, while at the same time creating new security challenges and being difficult to understand deeply.

Meridian Labs is a 501(c)(3) non-profit building an open source platform for understanding, evaluating, and testing AI models and agents. Our tools enable rigorous model testing and in-depth analysis of agent behavior.

Leading AI developers now routinely conduct pre-release evaluations to measure capabilities and assess risks, as illustrated by the detailed system cards published by organizations like OpenAI, Anthropic, and Google. In addition, model developers are collaborating on evaluations and security with government organizations like the US CAISI and UK AISI, and a robust group of indepenent third party evaluators is developing.

Our goal is to support this emerging ecosystem by creating a common platform shared by governments, academia, non-profits, and model developers. Our open-source software packages are in wide use across these organizations, and facilitate rigorous empirical work, deep insights, and cross-organizational collaboration.