In July, we shared that Debian developers were debating whether to ban AI-assisted contributions. Now, they’ve started voting on how to handle contributions made with LLMs and other generative AI tools, with the updated ballot now listing eight proposals and a “None of the above” option.
Voting began on August 15 and will stay open until August 28. Only Debian Developers can vote, using Debian’s ranked voting system to list their preferences among the options. Here’s a brief overview of what each of the nine options offers.
The first one is still the original proposal to ban LLM-assisted contributions by changing the Debian Social Contract. This would stop generative AI from being used for direct Debian work like packaging, software, documentation, translations, websites, and project communication. However, upstream projects made with AI help would not be affected.
The proposal gives several reasons for the ban, such as unclear copyright and licensing, doubts about the reliability of AI-generated work, extra review work for maintainers, aggressive scraping of Free Software resources, and the high resource use of large AI systems.
On the other hand, some proposals would allow AI-assisted contributions but hold the submitter fully responsible.
For example, the second option would clearly allow contributions that are partly or fully made by an LLM, as long as contributors check their technical quality, security, licensing, and usefulness, fully understand the changes, and disclose major AI help. It would also ban sending private or sensitive Debian information to untrusted external AI services.
A similar fourth option recognizes worries about generative AI but says banning it would be hard to enforce and not helpful. This proposal would accept AI-assisted Debian work if it follows the Debian Free Software Guidelines, is properly reviewed and understood by the contributor, and is marked as AI-assisted when needed.
The fifth proposal is even more open. It says Debian should not support or ban generative AI tools, but should apply the same standards for quality, correctness, maintainability, and legality to all contributions, no matter how they were made.
Contributors would still need to review, understand, test, and take full responsibility for AI-assisted work. They would be encouraged, but not required, to disclose AI use. The proposal also says that submitting generated material without proper human review does not fit Debian’s usual development practices.
Another option suggests a more careful approach. It would encourage contributors to avoid generative AI when possible and prefer human authorship and understanding, but it would not ban AI across the whole project. Disclosing AI use would be voluntary.
One unique proposal is called “Debian is created by humans.” Instead of banning AI tools for contributors, it focuses on what actually gets included in Debian.
With this approach, contributors could use generative AI for research, analysis, exploration, or critique, but could not submit AI-generated output directly as Debian packaging, patches, documentation, bug reports, project communications, or other Debian work. As the proposal says: humans create Debian.
Another, much stricter proposal asks Debian contributors to avoid LLM use as much as possible. It would require all Debian communication, like bug reports, mailing-list messages, Salsa discussions, and Planet Debian posts, to be written only by humans. Any AI use in Debian work would have to be disclosed, and violations could be handled under the project’s Code of Conduct.
The final, eighth proposal mainly looks at the environmental impact of large language models. It asks contributors to avoid using LLMs when possible, saying their resource use is a serious environmental issue. The proposal also admits that it is hard to reliably detect AI use and that Debian cannot realistically limit upstream projects whose software it distributes.
Voting ends on August 28. After that, Debian will have a clearer answer to a question many Free Software projects now face: not whether generative AI is used in software development, but how much it should be allowed to shape the work that becomes part of the project.
The “most open” option is the best option and the only one that should matter. Quality is the end all be all. If it’s bad quality, it doesn’t get accepted, if it is good quality it does. End of story.
I don’t understand why this is a debate. Let’s move past the age of ideological fake-activism please, it is getting very very tiring. Perhaps I’d understand if people were able to be nuanced. The witch hunts need to end, now.
love reading all these comnents here and everywhere, the very few all shilling for ai. read the language used = ridiculous, “the witchhunt needs to end now” is that you or rich elite tech billionaire loser or investors. these common pro-ai comnents are all PR bs. I just decided, where i did not care before ai, of course whether graphics, or coding ai can ve uses but fact, it only produces slop, which must (be used as a tool aka sloppy error prone base), unfoetunately those sloppy errors will wotk themselves in and humans use excuses and get lazy. fact. and when it comes to intricate detail as required in linux dev that risk is not acceptable. humans must edit and review ai, not the other way around. linux should be fully constructed by thoughful human construction, not alliwing slop and errors in period, then it should be peer reviewed. -Doctor X. But you will note the shill defender comnents say nothing, but ironicallyvempty general actual hypocritical ‘activist’ baloney. Guess what, I once used CentOS for my web server. Gee what happened several years ago planddmics? Next linux family big dog… Yep core father Debian. Strike 2. Notice a pattern fools?
Humans review the code for quality. If it’s bad quality (if it’s slop) it doesn’t get accepted. If it’s good quality, then it’s fine. That’s all I’m saying here. We don’t need philosophy for it because we can just be practical
I’m not a shill, I just don’t believe in being stuck in the past. I grew up believing in the future of AI well before it was the “thing” to hate on it, to be fashionable and trendy, and definitely well before it was an infinitely annoying culture war thing. People lose their minds over it, and can’t understand why it’s so annoying when they do that too.
When I made that comment I was very tired of the witch hunt junk all over the internet. Random people on the internet who use it to speed up their work are attacked by angry mobs or rabid politics-obsessed sorts of groups (because we really need more toxic ideology in this world…)
Now, it seems the decision has been made to mostly allow it so long as the code contribution is of good quality. AI is also getting better and better, so low quality submissions should become less and less. Which is great, the sooner denial-ism can die off and the sooner we can get consistent high quality results, the better. Then we can all move on
whatever they decide is ok with me.