LANGUAGES Signal 298
Rust Adopts Official Policy for AI-Generated Contributions
↗
Rust’s core teams now enforce a policy that limits how large-language-model output can be used in contributions to the main repository.
Engineers contributing to rust-lang/rust must treat AI assistance as a private aid, not a source of publishable code or text, and must disclose any AI involvement. Failure to follow the rules can lead to rejected pull requests, added review overhead, or temporary bans on AI-generated changes. The policy does not affect unrelated crates or projects that have not adopted it.
Written by elseif from the cluster below · every claim links back to a sourceThe three things worth knowing
AI may be used for personal assistance, but the final contribution must be authored and verified by the human submitter.
Generated text such as comments, documentation, or diagnostics cannot be submitted, even if edited afterwards, and must be clearly labeled if used.
If AI-generated changes exceed half of merged pull requests over a six-week window, new AI contributions are paused for at least ten days.
THE READ
What the cluster adds up to.
The Rust project has moved from a proposal stage to enforceable rules that restrict the role of large language models in contributions to its primary repository. The new guidelines apply only to the core teams that have formally adopted the policy, leaving other Rust repositories and independent crates untouched. This narrow scope means that most Rust developers outside those teams continue to operate under the previous, less formal expectations.
Contributors can still employ AI tools for personal learning, brainstorming, or private code checks, but they must retain full comprehension of the changes they submit. Any code or text that originates from an AI model must be disclosed, and the contributor must secure a reviewer who explicitly agrees to evaluate the AI-assisted work. The contribution is then marked with a specific label to allow the project to monitor its impact.
Publicly visible artifacts, such as issue comments, pull-request descriptions, documentation, and compiler diagnostics, cannot be derived from AI output, regardless of subsequent edits. This restriction prevents the project from inheriting low-quality or misleading content generated at scale, shifting the responsibility for correctness and clarity back to human reviewers. Reviewers retain the right to reject any submission that violates these constraints.
A safeguard limits the proportion of AI-generated pull requests: if they constitute more than half of merged changes over a six-week period, the project will temporarily suspend acceptance of new AI-assisted contributions. The pause lasts at least ten days and continues until the share falls below the threshold. This mechanism ensures that AI assistance does not overwhelm the review process or dominate the codebase.
For engineers working on Rust’s core, the practical impact is an added procedural step: disclose AI use, obtain explicit reviewer consent, and apply the ai-assisted label. Projects outside the affected teams are not bound by these rules, but developers should be aware that similar policies could be adopted elsewhere, requiring similar diligence.
Written by elseif from the cluster below · checked for specifics the sources never containedTHE CLUSTER