CourionAI
EN
Newsletter
← Glossary Company

METR

An independent group that tests advanced AI models for dangerous capabilities.

METR, short for Model Evaluation and Threat Research, is a non-profit that runs careful evaluations of frontier models to see what they can actually do, especially where that could be risky, such as autonomous hacking or self-improvement.

Its work matters because labs grading their own homework is not enough. Independent testing like METR’s is part of how the field tries to spot problems before a powerful model is widely released.