Debian votes to let contributors code with AI
The Debian Project has voted to permit contributors to use generative AI tools when creating software, packaging and documentation, though developers remain fully responsible for the quality of what they submit. Community members chose "Responsible Use of Generative AI" from a shortlist of eight competing proposals that ranged from an outright ban to environmental objections, reflecting a wider split within the open-source world over AI-assisted coding.
Under the winning proposal, Debian will "neither endorse nor prohibit" generative AI use, arguing the tools can boost volunteer productivity when used responsibly. It explicitly rejects "AI made a mistake" as an excuse for poor contributions, requiring all submissions to meet existing standards regardless of how they were produced, and states that blindly uploading unchecked AI output breaches Debian's development norms; disclosure of AI assistance is encouraged but not mandatory. Just under 600 people voted, with almost 450 valid ballots counted after rejections. The decision contrasts with Gentoo's outright AI ban and similar resistance from NetBSD and OpenBSD, while aligning Debian more closely with Linux's Linus Torvalds, who has welcomed AI contributions despite occasionally criticising AI-generated bug reports.
- Debian allows AI-assisted contributions but developers stay accountable for quality
- Winning proposal beat seven rivals, including a full ban
- Contrasts with Gentoo's AI ban; echoes Torvalds' pro-AI stance on Linux
New here? Start with this
Debian is one of the largest and oldest free operating system projects, and its code and tools underpin many other Linux distributions used on servers, desktops and devices worldwide. It's run by a community of volunteer developers who make decisions collectively, often through formal votes on project policy. This particular vote concerned whether contributors should be allowed to use generative AI tools, such as chatbots or code assistants, when writing software or documentation for Debian.
The question has divided open-source communities more broadly, with concerns ranging from code quality and legal uncertainty around AI-generated material to the environmental cost of running AI systems. Some projects, including Gentoo, NetBSD and OpenBSD, have chosen to ban AI-assisted contributions outright, while figures such as Linux creator Linus Torvalds have been more welcoming of the technology. Debian's members were asked to choose between several competing proposals reflecting these different viewpoints.
This matters because Debian's policy could influence how other open-source projects approach the same question, given its size and influence in the wider software ecosystem. It also touches on broader debates about accountability for AI-generated code, since any rules Debian sets affect who is responsible when mistakes make it into widely used software.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Allowing responsible AI use can help volunteers work more efficiently on code, packaging and documentation, potentially widening participation and easing maintenance burdens. Requiring every contribution to meet Debian’s existing standards keeps responsibility with the contributor, while optional disclosure avoids imposing a new process on work that is judged by its quality.
The case against
AI-generated work can introduce errors, insecure code or poorly supported changes that are difficult for reviewers to identify, adding to maintainers’ workload and potentially weakening trust. Critics may argue that disclosure should be required, and that a clear restriction would better protect Debian’s standards and volunteer capacity while the tools’ effects remain contested.