Five teams within the Rust project have adopted a formal policy governing the use of Large Language Models (LLMs) within the rust-lang/rust monorepo to address community challenges like increased review bandwidth demands and the erosion of effort-based signals in pull requests. The policy distinguishes between using LLMs to "answer, analyze, distill, refine, check, suggest, or review" versus using them to "create." While assistance is permitted, the use of LLMs for generating content is heavily restricted. Specifically, LLM-generated code is only allowed if it is pre-arranged, non-critical, high-quality, well-tested, and clearly disclosed. The policy mandates disclosure for machine translations, trivial changes, bug discovery, and when reviewing others' work. To maintain transparency, LLM involvement must be clearly marked in public documentation, PR descriptions, and GitHub comments. Reviewers are empowered to reject burdensome or non-compliant PRs and are not required to review LLM-generated code unless they volunteer. This framework aims to move away from inconsistent moderation toward a clear, actionable standard that prioritizes human understanding and community mentorship over mechanical, machine-generated contributions.