Debian GR 2024-003 adopts Choice 5: Responsible Use of Generative AI

The Debian Project published vote results confirming Choice 5 as the adopted policy on large language model usage. The text states that all contributions must meet identical standards of quality, correctness, maintainability, and legal compliance irrespective of production method. Contributor accountability is unchanged: each submitter must understand, review, test, and modify AI-generated material before inclusion. No disclosure mandate or tooling restriction was added.

Prior Debian positions on automated assistance, including copyright and licensing reviews from 2023, already required human verification of provenance. Choice 5 extends that pattern to generative models without introducing new enforcement mechanisms or audit trails. Comparable stances appear in the Linux kernel's documented reluctance to accept LLM-generated patches and in FreeBSD's informal contributor guidelines, both emphasizing unchanged liability rather than prohibition.

Operationally the policy shifts risk downstream: package maintainers now evaluate AI-derived code under the same scrutiny applied to human submissions, increasing review load where volume rises. Legal exposure remains with the human contributor for license violations or security defects. No central registry of AI-assisted uploads was created, leaving detection to existing code review processes.

Future impact depends on whether other distributions reference the Debian text when drafting their own rules. Within 18 months, measurable divergence in contribution patterns across projects citing Choice 5 versus stricter policies will indicate whether the responsible-use framing scales or fragments.