rust-lang/rust Is Adopting an LLM Policy

My take: Five teams in the Rust project, one of the most respected communities in open-source programming, today adopted a formal policy governing LLM use in contributions. The rule is clear and logical: AI can help you analyze, refine, or review, but the act of creating, whether documentation, code comments, compiler diagnostics, or the PR itself, must remain yours.

What stands out to me is not the restrictions themselves, but the safety valve they designed: if more than 50% of merged PRs in a six-week window are AI-generated, an automatic ten-day moratorium kicks in. It is a smart design because it acknowledges that the problem is not the tool, but volume and lack of judgment.

For those of us who use AI in our daily work, this decision confirms something we already know: the difference between using AI with intention and using it as a shortcut is enormous. AI makes you more efficient, but the judgment remains yours.

The question for any team, open-source or not: do you have a clear policy about where AI ends and your judgment begins?