LLM Policy for the Rust Compiler
Key point
A policy for LLM usage in contributions to the Rust compiler has been formalized.
Details
This is a policy document that lays out standards for LLM usage in contributions to rust-lang/rust.
This policy is designed to be kept as a living document on Rust Forge, and to be linked from CONTRIBUTING.md and the rustc/std-dev-guide. The target is the rust-lang/rust repository, and subtrees, submodules, crates.io dependencies, and other rust-lang repositories are out of scope.
The core aim is to create moderatable standards to reduce slop PRs. The document includes scope limitations such as the following:
- The long-term societal and economic impact of LLMs
- Environmental impact
- Copyright issues with LLM output
- Moral judgments directed at LLM users
It also asks that feedback focus on concrete policy improvement suggestions. The main review criteria are how a team's or individual's workflow is actually broken, and whether that problem is a blocking reason before merge.
As background, the document states that there are both people who find LLM-generated code and writing off-putting and people who see it as helpful for learning and exploration. At the same time, it explains that rust-lang/rust is already experiencing an influx of low-quality LLM-based PRs, so a clear policy is needed rather than case-by-case judgment.
As alternatives, a project-wide policy, differentiated standards per person, a looser permissive policy, and a stricter blanket ban were compared, and the current proposal is a compromise leaning toward greater moderation and understandability even if it blocks too much.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.