Debian developers have formally begun discussing whether to prohibit contributions created with large language models and other generative AI tools.
The debate started when Debian developer Matthias Geiger, with Jesse Rhodes, submitted a General Resolution proposal titled “Ban LLM contributions from Debian” on July 22, 2026. The proposal calls for Debian to explicitly reject direct contributions produced with or assisted by generative AI.
The proposed ban would apply to Debian source packages, packaging work, software developed by Debian, official web resources, documentation, translations, and project communications.
Importantly, the ban would not apply to upstream projects that use AI, software related to artificial intelligence, or patches and security fixes from upstream sources. In other words, Debian could still package third-party software developed with AI assistance, but Debian contributors would be prohibited from using such tools for direct project work.
Geiger’s proposal cites four main concerns: copyright, technical quality, impact on Debian’s volunteer community, and broader ethical issues related to generative AI.
Legally, the authors argue that the copyright and licensing status of AI-generated output is unclear. Moreover, the proposal questions whether AI-generated packaging and code can meet Debian’s quality standards.
The proposal also highlights community impact, citing that reviewing unreliable AI-generated submissions could increase maintainers’ workloads and contribute to burnout.
Ethical objections include the use of public content for AI training, disregard for licensing and robots.txt restrictions, the impact of aggressive scrapers on Debian infrastructure, and the computing resources consumed by AI systems.
To formalize the ban, the proposal would add a section to the Debian Social Contract stating that the project will not accept direct contributions created with or assisted by LLMs or other generative AI tools.
Coincidence or not, only two days later, a competing proposal submitted by Debian developer Lucas Nussbaum takes a considerably different approach. Rather than banning AI-assisted work, it would permit contributions that are partially or fully generated by an LLM, provided contributors follow a defined set of safeguards.
Under this alternative, contributors would remain fully responsible for the technical quality, security, usefulness, and licensing compliance of their submissions. They would also be expected to understand, explain, and defend their proposed changes.
In addition, significant AI assistance would need to be clearly disclosed. For software commits, contributors could use markers such as Generated-By or Assisted-By, with the exact form of disclosure left to the contributor.
The alternative proposal also prohibits sending confidential or sensitive Debian information to untrusted AI providers. This includes embargoed security reports, private project discussions, and other non-public material that could be exposed through external services.
The official discussion period for the General Resolution began on July 24, 2026. At this stage, Debian has not adopted either approach and has neither banned nor formally approved AI-assisted contributions.
Given Debian’s fundamental importance to the broader open-source community, the path the distribution ultimately chooses will carry significance far beyond the project itself. So, we will be following the debate closely and, as always, will keep you updated once a final decision is reached.
