This report documents the preparedness assessment of Code World Model (CWM), a model for code generation and reasoning about code from Meta. We conducted pre-release testing across domains identified in our Frontier AI Framework as potentially presenting catastrophic risks, and also evaluated the model's misaligned propensities. Our assessment found that CWM does not pose additional frontier risks beyond those present in the current AI ecosystem. We therefore release it as an open-weight model.
Cite
@article{arxiv.2605.00932,
title = {Code World Model Preparedness Report},
author = {Daniel Song and Peter Ney and Cristina Menghini and Faizan Ahmad and Aidan Boyd and Nathaniel Li and Ziwen Han and Jean-Christophe Testud and Saisuke Okabayashi and Maeve Ryan and Jinpeng Miao and Hamza Kwisaba and Felix Binder and Spencer Whitman and Jim Gust and Esteban Arcaute and Dhaval Kapil and Jacob Kahn and Ayaz Minhas and Tristan Goodman and Lauren Deason and Alexander Vaughan and Shengjia Zhao and Summer Yue},
journal= {arXiv preprint arXiv:2605.00932},
year = {2026}
}