OpenAI says it plans to let U.S. National Laboratories, the Department of Energy’s network of R&D labs, use its AI models for nuclear weapons security and other scientific projects.

Per CNBC, OpenAI will work with Microsoft, its lead investor, to deploy a model on the supercomputer at Los Alamos National Laboratory. The model will be a shared resource for scientists from Los Alamos, Lawrence Livermore, and Sandia National Labs, OpenAI says. It will be applied across a number of research programs.

One of those programs involves nuclear defense, controversially. OpenAI says it will support U.S. National Laboratories work focused on “reducing the risk of nuclear war and securing nuclear materials and weapons,” and will offer researchers with security clearances to consult on this.

“This is the beginning of a new era,” the company wrote in a blog post, “where AI will advance science, strengthen national security, and support U.S. government initiatives.”