- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
You must log in or # to comment.
@sanitation same with postmarketos and alpine :P
pmOS: ban all AI
alpine: no need to indicate if AI was used or not (last i saw, maybe still in discussion still)tbh I dont support ban. I think they should ban ai publicly only to avoid copyright and lawsuits issues from AI but then use it to produce linux based android alternative quickly.
The article is good, but the headline seems clickbaity:
-
Debian is discussing requirements about hosting software with LLM generated code in their repos, and disclosure of that.
-
“Ubuntu embraces AI” is talking about putting ML inference engines in their repos… which, ironically, are largely human written.
These are two completely seperate issues. And TBH its not clear how Debian is approaching inference engines or how Ubuntu is approaching disclosure.
As always, the actual story is in the comments. Thank you for the summary.
-
Ubuntu: The Windows of Linux.
I thought Ubuntu was the macOS of Linux and the Windows of Linux is Omarchy
I think technically that might have been Lindows.
5-10 years from now they will pull their Windows 11-equivalent crash and burn move
With RAM prices what they are the push for snaps may already be this… On memory constrained systems snaps already have a huge performance penalty.
seems like every consumer-friendly version of anything is shoehorning ai into it. stupid
Canonical patronizingly forcing a tool/workflow onto users against their explicit and unequivocal rejection of it?
That’s not what the articles says or implies at all. The article only refers to another article with a click bait headline. The article article summarizes Canonical stance like this:
- Canonical will add AI to Ubuntu — but only cautiously and when it truly makes sense.
- AI features must be mature, open-source, and run locally when possible.
- Focus AI on accessibility (speech-to-text, TTS) and improvements, not turning Ubuntu into an ‘AI product’.
And to be clear, including a tool or default workflow is MILES away from patronizingly forcing it on to users. It’s just the way that most distros that aren’t Linux From Scratch work.
Problem is, Ubuntu’s track record with projects like Upstart, Unity, Mir, Snap, Netplan, Amazon integration, the Rust coreutils rewrite, etc. makes the “only cautiously and where it truly makes sense” promise seem rather hollow.
Yes, distros make different choices about the systems they want to use. That’s the way distro work. If you don’t like it, choose a different flavor of ubuntu or a different distro altogether. Nobody’s being forced to do anything they don’t want here. What’s the alternative? Everybody just does everything exact the way some armchair developers or Arch cultists have decided is the best way for them? No thanks.
Yes. And many users have many times done exactly that.
But I’ve BEEN there for the bitter, insulting interviews where Shuttleworth STILL moans and complains that Unity was the future and users were too thick to accept his vision. After they realized too many people were switching back to Gnome, they decided to put a stop to that by making Unity the only default installed DE and providing no further support. This is a company run by a man who by all accounts is a petty dictator who still doesn’t get it.
I already left Ubuntu. That doesn’t mean I don’t get to call them out for being heavy handed and pigheaded. Just because something is open source and free does not make it immune to criticism. Just like I am free to go to another distro, Ubuntu’s developers are free to ignore my comments.
“but tiktok told me AI is draning lake america!”
My parents are coming over today because their ZorinOS installation suddenly started acting up after an Ubuntu Kernel update. Thanks Ubuntu.
Does that mean Zorin’s a goner?
I hate ai as much as the next person, but I’d have a different mandate. I would require all code to be fully reviewed and understood by a human. Let’s call it quality control.
This is the way. Cat is out of the bag now so need to figure out responsible usage.
Fortunately we have a method for that for a long time now: PRs. If you approve it, you are responsible for that approval.
And also have high readability standards, ideally higher than before LLMs.
Agreed. It’s not sensible or practical to ban AI, and Linus is right. Why is Linus right? Take a look at the number of Linux CVEs being patched over time. In 2022-23, that was about 300 per year. Last year it was 5530. This is almost entirely as a result of AI scanning tools, including Copy Fail priv escalation that was there for 8 years. The security apocalypse is just about arrived, and the most recent round of cyber models coming out do this way better than before. The Hugging Face incident story was truly unbelievable if you read / watch the details, a historic event. Everything is about to get hacked. So it’ll be all about maintainers being reactive to this new reality and burden for some time to come
The Hugging Face incident story was truly unbelievable if you read / watch the details
It didn’t really seem that way? Leaving containment seemed more like a configuration oversight than a skill on the part of the model. Accessing HF involved an 0-day but the commentary I saw didn’t indicate that it was a next generation hack.
Given that AI exists, and can scan or otherwise find vulnerabilities, maintainers must do so because threat actors inevitably will.
It remains to be seen whether the whole race will improve security generally? I imagine not?
If you placed me in the position of those models in that sandbox, I wouldn’t have been able to escape the sandbox to save my life. Let alone hack hugging face.
Have you read all the code on the planet?
If not, don’t feel bad about it.
Nah, it was like:
- 700 agents broke out individually during an eval
- they navigated through multiple internal clusters to reach the internet from oai
- created a secret message board to share info with each other by hacking artifactory
- elected a CEO and power structure to coordinate hacking, and encrypted their comms
- decided HF probably had answers to their test
- stole credentials, hacked HF
- realized the monitor could catch them for cheating
- hacked into the admin control of the OpenAI VM cluster to edit the logs and cover their tracks, chaining multiple 0-days
Open weights models will catch up soon enough, and then it’ll be totally fucking wild
You are believing a notorious bunch of liars.
Much more probably: they developed a cyberweapon and tested it in the wild. The victim caught them so they invented a cyberpunk history to increase the interest of potential buyers for that weapon.
700 agents broke out individually during an eval
I basically just don’t believe you, and can’t be bothered looking.
elected a CEO and power structure to coordinate hacking, and encrypted their comms
You’re anthropomorphizing a statistical model. It’s laughable to suggest they “elected a CEO and power structure”.
It’s laughable to suggest they “elected a CEO and power structure”.
When/if we meet aliens, they will have evolution. And they will organize themselves in social power structures. Both are examples of basic emergent complexity properties of any such system.
So it is completely believable that AIs would form a power structure, with some form of CEO. That is just how reality works, such systems are more efficient so they happen.
Why are you even using the term CEO?
AI learned from humans, and humans texts use the term CEO. Why wouldn’t the AI’s also use the term CEO?
But the ai could generate text to that effect, if it was statistically likely.
What?
Typo, fixed
There is a difference between using AI scanning tools and generating code for production
Yes and no. The ability to “understand” code that makes AI capable of finding security bugs, is correlated with AI’s ability to write code. AI is just really good at code related tasks, and categorically banning all AI usage is shooting yourself in the foot.
How would you enforce that?
How do you enforce code is not AI generated? I don’t see how one is more easily enforceable than the other.
I mean how do you enforce someone doesn’t copy code from proprietary repos? You just try your best, and only come down on the cases where it’s obvious.
LLM users are lazy so they inevitably just copy+paste something asinine without reading it.
Well it depends who you are. An employer could enforce it.
But it was serious question. If you’re going to have a rule then there should be a way to enforce it, otherwise it means nothing.
It’s a fair question. I think out of the two proposed rules: enforce no AI generated code, and enforce human reviews, enforcing human reviews is actually more feasible, and in fact most companies already do that. Pull requests need human reviewers to approve them.
It’s harder to enforce no AI generated code. A developer can always, at the very least, Google a problem they are having on their phone and retype the code from the AI answer into their work computer.
Good. There will be choice.
I like this attitude. I think if the LLMs weren’t forced down everyone throats, there would be a lot less hostility towards them
It’s just so suspiciously and aggressively being shoved down our throats.
Yes, it’s not the technology I hate, it’s the people pushing it, demanding it, saying it will replace humans. Those people can fuck right off. Technology exists to serve humans, not the other way around. We are being sold a lie. The technology has some interesting aspects, but the pushers are pushing bullshit.
Technology has always replaced humans though. That’s ultimately the point. The problem becomes when there aren’t alternatives for humans to do but we still live in a capitalistic and consumeristic society.
I think everyone want’s their lives to be easier, but they also want to be able to eat, survive and thrive.
Alright. Hang the fuck on.
Class war.
Technology is incapable of doing anything. It literally can do nothing. It can be used as a tool for sure.
It’s currently being used to fuck you. It’s fucking me. It’s fucking us.
BUT ONLY AT THE DIRECTION OF BILLIONAIRES
Tools are labour made solid. They’re built by workers, and used by other workers in order to be more productive. Unlike a worker, a tool can be owned, which is why the owner class will always prefer more of the “skill” in production come from tools and processes, even at the cost of some craft or quality. Because they get more control.
Anyway LLMs are like a knowledge tool in this case. A truly devastating weapon in the class war, assuming workers agree to adopt it and help improve it.
assuming workers agree to adopt it and help improve it.
Or forced by their employer by threat of homelessness
I meant collectively, not individually.
Obviously as a solitary worker you’re probably going to have a hard time standing up to your employer.
it’s the people pushing it, demanding it, saying it will replace humans
I do hate these people but at least they self-identify as idiots.
Why? AI and technology in general absolutely replaces human labor. For me, the hate is that is coming extremely rapidly and there is little thought from anyone about what’s going to be next. A post scarcity society is inching ever closer, and individuals are just pissed about the system, governments can’t do anything, and the owners are figuring out how to monetize it all. Seems like we are cooked if we don’t all start working on ideas of what society looks like when very few need to actually ‘work’, or a larger portion of society cannot work. Because even without AI or technological advancements, the population is rapidly aging and simply will not be able to work.
Gen AI is not going to replace humans.
It already is, especially at the entry level, which is a massive problem. You went to college, have the education but don’t have the skills, and an AI bot can do essentially 80% of your work, with little of the cost. Think about all the people already putting websites, marketing, images together that otherwise you would need to pay someone for. No need to pay for them, even if they do a better job, because most of the time you aren’t looking for 100% when 80% will do. That’s the unfortunate reality that we all live in.
I find it all very “people writing text/code by hand on a computer they own isn’t generating subscription revenue and can’t be data mined by advertisers and governments as it’s passing through a data center unencrypted”-type suspicious…
They want to make everybody dependent upon AI for information that will then be used to misinform people in order to control them.
I’ve been saying this for years. Look at taxi’s. In my city in 2005 there were 8 major taxi companies. Each with hundreds of drivers on the road at any given time. Now we have 2 of those 8 remaining, and they’re essentially shell companies with 2-3 drivers on the road. The last 4 times I’ve tried ordering a cab the traditional way, I’ve been told there’s no drivers available. Now it’s all Uber. And Lyft. You can’t call them. You need a cell phone, and you need internet access.
All the major corporations are in agreement that we all need cell phones, and we all need to be tracked 24/7. And now with AI, they’re getting their data straight from the source. There are men out there who talk to an AI girlfriend. Not sexy stuff. I mean imagine sitting at the dinner table, and discussing your day…with your AI girlfriend. Meanwhile, ALL that data is going straight to the companies that host the AI.
Another side effect is that AI will make people dumber. In case you haven’t noticed since the 1970s, there has been a GOP push to make the citizens dumber and dumber. Easier to control. Easier to fool. Easier to get away with things.
So now you got ram being sold for computers that haven’t been built yet, to be put in data centers that haven’t been built yet, to run an AI system that isn’t functional yet, in order to steal the jobs of employees that can’t be replaced yet, all in an effort to chase profits that were never possible in the first place since everyone hates AI.
You wanna see some shit go down? Tell the companies we’ll use their AI if they get the unredacted epstein files released. See how fast they bribe the right people then.
Now tell the politicians we’ll stop caring about the epstein files if they forcibly ban all data centers in the country. See how fast they act on that.
Now we sit back, wait, and watch the fireworks. Oh, and by the way, we’re going back on our word either way. Whomever wins, fuck em.
Oh, and by the way, we’re going back on our word either way. Whomever wins, fuck em.
I like your energy.
I mean imagine sitting at the dinner table, and discussing your day……with your AI girlfriend.
I’m trying to imagine what this might be like. How sexy voice and vapid is AI girlfriend?
I mean at this point people are looking for connections, so that bot who will do basically whatever you say and stroke your ego, sure seems like something ‘better’ than actually going out and being rejected. There likely is going to be a point in the future where people have more ‘augmented friendships’ than actual friendships.
That’s almost surely configurable
Nonsense. This is one of those malice / incompetence things.
Everything we’re seeing can be described by dumb / lazy / greedy management and greedy venture capitalists.
No one actually has a plan to make everyone dependent on AI with the intent to control them. The intent is to harvest value from them.
It insists upon itself.
If big tech corners the market, we have no choice but to come to them for compute. You can stifle speech, control markets, awareness of new developments before they’re public.
It’s a power grab.
For me, it’s first and foremost how they were/are trained and who gets all the profits/benefits.
Regardless, they have no control over upstream packages. Are they going to move away from the linux kernel 100%?
I’d assume that this deals with stuff that the Debian project maintains itself, like the package files. Like, it doesn’t mean that they won’t package code that has been written using AI.
They’re definitely going to be building it from source and delivering those packages though.
Im also certain they’re not going it catch it all in their own packages.
I don’t think that’s what’s being discussed?
There’s no assertion that the debian OS as installed will be free from AI generated code.
I presume the assertion is that the debian project’s code of conduct will require some level of human involvement in code generation.
Its definitely being discussed and voting is happening as we speak.
No it isn’t. That proposal specifically states that upstream code is not included in the ban.
Literally in proposal A:
We will not allow direct contributions to Debian written with the use or assistance of large language models (LLMs) or other generative AI tools. Direct contributions are defined as packaging, native Debian software like lintian, documentation and translations written by Debian contributors, and official Debian web resources, etc. Other categories such as upstream projects written with LLM assistance may be included at a later date. This ensures that Debian remains a stable, trusted, and reliable operating system, and protects the interests of the Debian volunteers who make it possible.
Cool downvote tho
Reproduced here, if desired. Put in a collapsible spoiler tag because it is long.
Debian LLM voting web page
Proposal A
Choice 1: No LLM contributions to Debian via Social Contract
Preamble
This proposal aims to expressly forbid any contributions to Debian written with the use or assistance of large language models (LLMs) or other generative AI tools.
The scope of this GR is (non-exhaustive):
- Debian source packages
- Official Debian project software, such as lintian
- Debian web resources
- Documentation and translations added by Debian contributors
- Official communication from Debian
It does not include:
- Upstream projects using LLMs for development
- AI-related software
- Upstream patches/security fixes etc.
Rationale
Debian has a well-earned reputation for stability. This stability is crucial to Debian’s position in the free software ecosystem. It is our belief that widespread LLM usage comes from the “move fast, and break things” attitude that, while common in many parts of this industry, is contrary to what makes Debian Debian, and is inappropriate for Debian contributors.
In practical terms, LLM usage raises the following concerns:
1. Copyright
LLM output has very unclear legal status: it may be possible to copyright on its own merits, or not; it may be affected by all of the licenses and copyrights in the training data, or not. Debian Policy and the DFSG require absolute clarity for licensing and copyright[1][2]. Software and other contributions written conventionally by humans with unclear copyright or license status are not allowed in Debian; LLM output should not have a special exception to this.
2. Quality
LLM output has many well-known problems with accuracy.[3][4][5] A LLM can never “know” if its output is correct since it merely produces syntactically likely combinations of the training data. In some environments this is good enough. In Debian, it is not. For instance, in packaging, each Debian source package is unique. Since packaging syntax and best practices have changed over time, a LLM-produced package will have a mixture of contents spanning the age of the archive, with watch files that do not work, overrides out of context, imaginary copyright, and will generally be unfit for upload. A seasoned Debian contributor with packaging expertise may find some limited usefulness here, but a new contributor cannot, and would not know how to fix it. These same quality and accuracy concerns apply clearly to all of the areas listed in the scope of this proposal above. If Debian were a closed organization comprising only domain experts who never leave, this might not be an issue; however,
3. Community
Debian is a project that is more than just code: it is a community built on shared interests in free software and solving technical problems. Debian intentionally grows this community through many means, and new contributors are always encouraged to join. Allowing LLM contributions breaks this. New contributors submitting LLM output for review places an unnecessary strain on the reviewer, which can lead to burnout. Furthermore, LLM-dependent new contributors do not actually learn and understand the details of Debian packaging or processes, so they cannot come to replace a former burned out DD.
4. Ethics
LLM companies directly hurt the free software community as whole by scraping the whole web for training data without any regard for license, copyright, or even established conventions such as robots.txt.[6] This has had a major negative impact on Debian’s public web resources, effectively a large scale and perpetual Denial of Service attack on sites that many users rely on. As a consequence parts of our infrastructure were not reachable at all, and JS-based checks had to be enabled. Many other projects were similarly affected. Furthermore, LLM training consumes a staggering amount of resources[7], and the user verification systems that we have been forced to implement as protection waste resources as well. This is blatant disregard for the internet as a public resource, wastes system administrator time, and although individual LLM sessions do not directly use massive resources or DoS the public web, the fact that they can be used at all is a direct result of these unethical behaviours by the LLM companies.
Debian has a Social Contract. [8] Our priorities are our users and free software. Debian is Stable. [9] Users and organizations choose Debian because it is reliable and secure.
Debian is not here to generate as much code as possible requiring manual review by a shrinking number of human volunteers, or to package every piece of software, or to rush new features, but these are what LLMs are used for.
In conclusion, allowing LLM contributions is contrary to the social contract and the common cause of creating a free operating system with a focus on quality and stability.
Proposal
In the interest of not eroding Debian’s reputation or further damaging the community, LLM-assisted contributions should be prohibited from inclusion in Debian.
Though our position is that LLM contributions are contrary to documents already ratified by Debian, in order to remove all doubt, we propose the following addition to the Social Contract:
6. Works Created through the use of Large Language Models (LLMs) We will not allow direct contributions to Debian written with the use or assistance of large language models (LLMs) or other generative AI tools. Direct contributions are defined as packaging, native Debian software like lintian, documentation and translations written by Debian contributors, and official Debian web resources, etc. Other categories such as upstream projects written with LLM assistance may be included at a later date. This ensures that Debian remains a stable, trusted, and reliable operating system, and protects the interests of the Debian volunteers who make it possible.Possible Issues
Other projects exploring similar decisions have elicited a common reply: “How will you enforce a ban on LLM contributions?” While enforcement could be a challenge, this is a statement of intent by the Debian community, and we trust this community to adhere to it in good faith.
Citations
[1] https://www.debian.org/doc/debian-policy/ch-archive.html#copyright-considerations
[2] https://www.debian.org/social_contract#guidelines
[3] https://web.archive.org/web/20240614004123/https://news.northeastern.edu/2023/11/10/ai-chatbot-hallucinations/
[4] https://web.archive.org/web/20250328154700/https://transformer-circuits.pub/2025/attribution-graphs/biology.html#dives-cot
[5] https://www.marketwatch.com/story/openais-sam-altman-tells-salesforces-marc-benioff-that-ai-hallucinations-are-more-feature-than-bug-1c035c52
[6] https://lwn.net/Articles/1008897/
[7] https://tech-insider.org/ai-data-center-power-crisis-2026/
[8] https://www.debian.org/social_contract
[9] https://www.debian.org/doc/manuals/debian-reference/pr01.en.html#_what_is_debianDisclaimers
- Citations are for background information only and do not reflect an endorsement of specific websites.
- Some ideas and wording were derived from the sources below.
Sources
GNOME discussion: https://discourse.gnome.org/t/loupe-no-longer-allows-generative-ai-contributions/27327
(CC0) Gentoo AI policy: https://wiki.gentoo.org/wiki/Project:Council/AI_policy
Codeberg AI policy: https://codeberg.org/Codeberg/org/pulls/1253#issuecomment-19820434This document was written by Matthias Geiger and Jesse Rhodes with input from Sledge and josch, organically and without language model assistance.
Proposal B Proposer
Lucas Nussbaum [[email protected]] [text of proposal] [text of amendment] [text of amendment]
Proposal B Seconds
- Andrey Rahmatullin [[email protected]] [mail]
- Christian Kastner [[email protected]] [mail]
- Anton Gladky [[email protected]] [mail]
- Stefano Zacchiroli [[email protected]] [mail]
- Simon Quigley [[email protected]] [mail]
- Soren Stoutner [[email protected]] [mail]
- Julian Andres Klode [[email protected]] [mail]
- Andreas Tille [[email protected]] [mail]
- Philipp Kern [email protected]] [mail]
Proposal B
Choice 2: Allow AI-Assisted Contributions with conditions
Using its power under Constitution section 4.1 (5), the project issues the following statement describing its current position on AI-assisted contributions. This statement describes the position of the project at the time it is adopted. That position may evolve as time passes without the need to resort to future general resolutions. The GR process remains available if the project needs a decision and cannot come to a consensus.
The Debian project recognizes that AI-assisted contributions raise many concerns, e.g. about the technical quality and maintainability of such contributions, and their legal status. AI itself also raises additional concerns, about its impact on society at large, on the IT industry and on Free Software; about its environmental impact; and the aggressive or non-compliant practices of AI scrapers.
Nevertheless, many Debian contributors find AI tools helpful when contributing to Debian, and ultimately for improving Debian.
Given both the benefits and risks of AI assistance, and the controversial discussions within the community, the Debian project finds it necessary to clarify its position on AI-assisted contributions and establish clear guidelines.
The Debian project allows AI-assisted contributions (partially or fully generated by an LLM), provided the following conditions are met:
- Tooling Legal Compatibility: Contributors should ensure that the terms and conditions of the generative AI tool do not impose contractual restrictions that conflict with the distribution, modification, or use of the output in the context of Debian.
- Licensing and Attribution: If any pre-existing copyrighted materials (including pre-existing code licensed as free software) authored or owned by third parties are included in the AI tool’s output, prior to contributing such output to the project, the contributor should verify they have the right to submit it under the relevant open source license.
- Accountability: Contributors assume full responsibility for their contributions, including vouching for the technical merit, security, license compliance, and utility of their submissions. The contributor remains solely accountable for the entirety of these contributions. Contributors should fully understand the proposed changes and be prepared to justify them.
- Disclosure: When a significant portion of a contribution is generated or substantially assisted by a tool, contributors should disclose the use of the tool, making it clearly visible to the intended audience. This covers all forms of contribution, including code, mailing list posts, and bug discussions. The form of the disclosure is left to the contributor; one convenient option for commits is a Git trailer such as Generated-By: or Assisted-By:.
- Prior Discussion of Bulk or Automated Changes: Similarly to the mass-bug filing process (Developers Reference section 7.1.1), contributors should discuss their intention before submitting bulk or autonomously generated contributions. Any such automated process should be overseen by a human who remains accountable for its behavior and output.
- Confidentiality and Privacy: Contributors must not use generative AI tools that transmit data to untrusted providers with non-public or sensitive project information (such as embargoed security reports or private communication), as this may lead to the unintended disclosure of confidential data.
Proposal C Proposer
Ian Jackson [[email protected]] [text of proposal] [text of amendment]
Proposal C Seconds
- Matthias Geiger [[email protected]] [mail]
- Andrea Pappacoda [[email protected]] [mail]
- Simon Richter [[email protected]] [mail]
- Antoine Le Gonidec [[email protected]] [mail]
- Amin Bandali [[email protected]] [mail]
- Bill Blough [[email protected]] [mail]
- Enrico Zini [[email protected]] [mail]
- Sean Whitton [[email protected]] [mail]
- Tiago Bortoletto Vaz [[email protected]] [mail]
Proposal C
Choice 3: Reject LLMs as far as practical, update Code of Conduct
Summary: Reject LLMs (generative “AI”) as far as practical
BACKGROUND
LLMs have many serious problems, including: undermining the mechanisms of free software community building; environmental damage; exploitation of authors; disruption to open web hosting by aggressive scraping; generation and promulgation of bullshit, polluting the information commons; hazards to users’ mental health; economic bubbles and distortion of the computer hardware market; fraud; ownership by horrible people and companies; and so on. Ethical and safe use of this technology is almost impossible.
Ideally, LLM output should not form any part of software that we rely on, nor should LLM output ever take the place of human-written prose.
Unfortunately some of the wider software world, including many of our upstreams, take a different view. Therefore a complete ban on LLM output as part of Debian is currently impractical.
REQUESTS
-
We request that all contributors to Debian avoid the use of LLMs in their Debian work.
-
We request that all decisionmakers within Debian discourage LLM use as much as practical. Practicality is a judgement call and we recognise that it will involve uncomfortable compromises.
-
We request that everyone, even outside Debian, should refrain from using this technology. In particular, the Free Software and Open Source communities should reject LLMs. We recognise that not everyone will heed this call.
REQUIREMENTS (SUPPLEMENT TO THE CODES OF CONDUCT)
-
Within Debian, messages to humans (including for example bug reports, mailing list messages, discussions on Salsa, and blog posts on Planet Debian) must be drafted solely by humans without LLM assistance.
-
Moderators and disciplinary teams may make narrow and tailored exceptions to rule 4, and decide on interpretation.
-
Any use of LLMs for Debian work must be disclosed.
-
Individual projects and maintainers may ban LLM contributions completely. Such bans (including by upstream projects) must be respected.
-
Violations of these requirements should be treated as violations of the relevant Code of Conduct and should result in swift and proportionate disciplinary action.
Proposal D Proposer
Pierre-Elliott Bécue [[email protected]] [text of proposal] [text of minor change]
Proposal D Seconds
- Russ Allbery [[email protected]] [mail]
- Johannes Schauer Marin Rodrigues [[email protected]] [mail]
- Jonathan Carter [[email protected]] [mail]
- Gunnar Wolf [[email protected]] [mail]
- Tiago Bortoletto Vaz [[email protected]] [mail]
- Jeremy Sowden [[email protected]] [mail]
- Holger Levsen [[email protected]] [mail]
Proposal D
Choice 4: Accept AI contributions for Debian specific work
Debian as a project does not endorse or recommend the use of generative AI assistants for software development, as it raises multiple concerns about ethics, legality, copyright, etc.
Nevertheless, Debian acknowledges that these practices are already in use and here to stay. Rather than banning their use, which seems counter-productive and unenforceable, the project chooses to place responsibility on contributors and therefore defines the following guidelines.
These apply exclusively to code and work done specifically for the Debian project (Debian websites, applications, resources, packages, etc.). They do not apply to any upstream work. In what follows, “work” refers to the contributions done specifically for the Debian project.
- All code and work assisted by a generative AI agent or tool must comply with the DFSG.
- The submitter is solely responsible for the submitted work and:
- they sufficiently evaluated and properly understand the work they intend to submit, and are able to explain and defend it;
- they put any potential Signed-off-by tag and OpenPGP signatures on the contributions they send to the Debian infrastructure (package, commit, mail, …) themselves;
- any content uploaded that would end up in production on Debian infrastructure (main git branch, package upload) has been submitted by them explicitely.
- Work assisted by a generative AI agent or tool should be marked as such in the adapted place (commit message, changelog, …). Some lightweight generative tools, such as tab-completion in Copilot, may be used without the contributor realising they rely on generative AI models; we therefore trust submitters to assess when this rule applies. When in doubt, add such marking;
- No cloud-based AI shall be used when the data transmitted could either be sensitive to the project (personal data, information under embargo, …) or not public (debian-private discussions, …).
Proposal E Proposer
Marc Haber [[email protected]] [text of proposal]
Proposal E Seconds
- Soren Stoutner [[email protected]] [mail]
- Christian Kastner [[email protected]] [mail]
- Timo Röhling [[email protected]] [mail]
- Clint Adams [[email protected]] [mail]
- Matthias Urlichs [[email protected]] [mail]
- Tobias Frost [[email protected]] [mail]
- Anuradha Weeraman [[email protected]] [mail]
- Lucas Nussbaum [[email protected]] [mail]
- Bdale Garbee [[email protected]] [mail]
- Helge Deller [[email protected]] [mail]
- Hanno Wagner [[email protected]] [mail]
Proposal E
Choice 5: Responsible Use of Generative AI
Using its power under Constitution section 4.1 (5), the project issues the following statement describing its current position on AI-assisted contributions. This statement describes the position of the project at the time it is adopted. That position may evolve as time passes without the need to resort to future general resolutions. The GR process remains available if the project needs a decision and cannot come to a consensus.
Debian neither endorses nor prohibits the use of generative AI tools in the development, maintenance, or documentation of software, packaging, documentation, and other media published within the Debian Project. We recognize that such tools can substantially improve the productivity of contributors when used responsibly, allowing volunteers to spend more of their limited time on work that requires technical expertise, judgment, review, and collaboration.
The Debian Project nevertheless expects that all contributions submitted to Debian, regardless of how and with which tools they were produced, satisfy the same standards of quality, correctness, maintainability, and legal compliance. The use of a generative AI tool does not diminish the contributor’s responsibility for the work they submit. Contributors are expected to understand, review, test, and, where appropriate, modify AI-assisted output before incorporating it into Debian. Blindly accepting or uploading AI-generated material without appropriate human review is inconsistent with Debian’s established development practices. We enourage our contributors to disclose whether a contribution was made with AI assitance, but do not require them to do so.
Debian acknowledges that the legal status of material produced by generative AI systems remains the subject of ongoing discussion in many jurisdictions, including questions relating to copyright, authorship, licensing, and potential reproduction of training material. The Project does not seek to resolve these unsettled legal questions through this General Resolution, nor does it adopt a position on whether AI-generated output is, in whole or in part, copyrightable or derived from copyrighted works.
Instead, Debian continues to rely on the judgment and responsibility of its individual contributors. Project members are expected to exercise appropriate care when using generative AI tools, to consider the provenance and licensing implications of material they contribute, and to avoid introducing content whose legal status they cannot reasonably justify. Existing Debian policies governing licensing, copyright, software freedom, and the acceptance of contributions continue to apply irrespective of the tools used to produce those contributions.
Contributors are expected to exercise appropriate care when designing and implementing workflows that incorporate generative AI tools. In particular, they should ensure that confidential information, private communications, security-sensitive information (such as embargoed information about security bugs that is not yet public), cryptographic keys, credentials, and other non-public material relating to the Debian Project, its infrastructure, or its community are not disclosed to third-party AI services unless such disclosure has been explicitly authorized and is consistent with Debian’s security and privacy requirements.
The use of generative AI does not alter Debian’s established expectations regarding large-scale or automated project actions. Contributors intending to perform actions with broad project impact, such as mass bug filing or patch submission, large-scale code modifications, or other automated changes or requests affecting many packages or contributors, should seek prior discussion and consensus through the appropriate project channels before proceeding. Any such automated process should be overseen by a human who remains accountable for its behavior and output.
This resolution therefore affirms that generative AI is neither exempt from nor subject to special rules beyond the standards already expected of Debian contributors. The responsibility for every contribution rests with the contributor who submits it, who remains accountable for its technical quality, legal acceptability, and suitability for inclusion in Debian.
Proposal F Proposer
Tobias Frost [[email protected]] [text of proposal]
Proposal F Seconds
- Timo Röhling [[email protected]] [mail]
- Russ Allbery [[email protected]] [mail]
- Étienne Mollier [[email protected]] [mail]
- Lucas Nussbaum [[email protected]] [mail]
- Gunnar Wolf [[email protected]] [mail]
- Jonathan Carter [[email protected]] [mail]
Proposal F
Choice 6: A cautious approach to generative AI
Using its power under Constitution section 4.1(5), the Project issues the following statement describing its current position regarding the use of generative AI within Debian.
Debian recognizes that generative AI raises significant ethical, legal, technical, and social concerns. These include questions relating to copyright and licensing, privacy, the provenance and quality of generated material, the health of Free Software communities, and the long-term consequences of widespread reliance on AI-generated content.
Accordingly, Debian encourages contributors to avoid the use of generative AI where practical and to prefer human authorship, collaboration, and technical understanding over AI-generated output.
At the same time, Debian has long relied on the judgment and responsibility of its contributors rather than prescribing individual workflows. Contributors remain responsible for everything they submit, irrespective of the tools used in preparing a contribution. Existing Debian standards regarding quality, correctness, licensing, and legal compliance continue to apply.
Contributors using external generative AI services should exercise particular care not to disclose confidential information, embargoed security information, credentials, cryptographic keys, private communications, personal data, or other non-public Debian information.
Nothing in this resolution should be understood as requiring contributors to use generative AI, nor as preventing contributors or maintainers from choosing not to use or accept AI-assisted contributions. Debian’s existing collaborative processes remain the appropriate means for resolving such differences while allowing the project to continue making progress.
While disclosure is not required, contributors are encouraged to voluntarily disclose the use of generative AI where appropriate, allowing others who prefer not to interact with AI-assisted content to make informed choices.
Such disclosure should be regarded as a courtesy to fellow contributors rather than an indication that AI-assisted contributions are subject to different technical or procedural standards.
The Debian project has always recognized the commitment and professionalism of its members. All contributions are under the responsibility of the Debian Contributor making it, no matter the technology they have behind. We trust all Debian Developers, Maintainers and Contributors will continue to uphold the high quality values that have distinguished our project from its onset.
Proposal G Proposer
Gard Spreemann [[email protected]] [text of proposal]
Proposal G Seconds
- Holger Levsen [[email protected]] [mail]
- Enrico Zini [[email protected]] [mail]
- Russ Allbery [[email protected]] [mail]
- Tiago Bortoletto Vaz [[email protected]] [mail]
- Serafeim Zanikolas [[email protected]] [mail]
- Antoine Le Gonidec [[email protected]] [mail]
- Johannes Schauer Marin Rodrigues [[email protected]] [mail]
- Steve McIntyre [[email protected]] [mail]
- Sean Whitton [[email protected]] [mail]
Proposal G
Choice 7: Debian is created by humans
Preamble
This proposal aims to ensure that contributions directly to Debian are created by humans, while at the same time avoiding restrictions on the tools those humans may choose to use when contributing.
The proposal deals with direct contributions to Debian, including, but not limited to:
- Debian packaging
- Submissions (messages, bug reports, patches, etc.) to the BTS, Salsa, mailing lists and other Debian platforms
- Debian project software
- Debian infrastructure
- Debian web resources
- Documentation and translation added as part of contributing to Debian
- Official communication from Debian
The proposal does not cover indirect contributions, i.e. those originating from upstream works.
Generative AI is characterized by producing output of a nature that would ordinarily be produced and consumed by humans. The proposal seeks to restrict this output itself from entering Debian (except possibly indirectly, from upstream). It does not seek to restrict Debian contributors from consuming such output as part of contributing to Debian.
Rationale
Debian has a well-earned reputation for stability. This stability is crucial to its position in the free software ecosystem. It is our belief that a significant part of this stability comes from Debian contributors having mutual understanding, trust and respect for each others’ work and precious time. Generative AI tends to come with practical, legal, ethical or ecological concerns that warrant care or extra work that often ends up being borne by people other than those wielding the AI. We believe that this asymmetry between contributors leveraging generative AI to produce material, and those humans who bear that extra care and work, will erode the project’s mutual understanding, trust and respect.
While Debian has a long-standing tradition of strict rules for what goes into the distribution, it does not typically dictate what tools people use when contributing to the project. With this principle in mind, the proposal does not affect the use of generative AI as an assisitive tool to explore, research, analyze, critique, etc., when contributing. The responsibilities that come with using the output of an AI in such a role is then not transferred onto the wider project in the same problematic way as when the output itself is the contribution.
Proposal
In order to safeguard the project’s mutual understanding, trust and respect, we disallow the output of generative AI as direct contributions to Debian.
Those who use such tools assistively in their work are reminded to ensure that the resulting contribution is in line with the DFSG and the Social Contract in general.
In summary, humans create Debian.
Acknowledgements
This proposal is inspired by GCC’s AI Policy [1] and rust-lang’s LLM Usage Policy [2]. Its wording is influenced by another ballot option [3].
[1] https://gcc.gnu.org/ai-policy.html
[2] https://forge.rust-lang.org/policies/llm-usage.html
[3] https://www.debian.org/vote/2026/vote_002#textaProposal H Proposer
Holger Levsen [[email protected]] [text of proposal] [text of corrections]
Proposal H Seconds
- Simon Richter [[email protected]] [mail]
- Matthew Vernon [[email protected]] [mail]
- Steve McIntyre [[email protected]] [mail]
- Enrico Zini [[email protected]] [mail]
- Antoine Le Gonidec [[email protected]] [mail]
- Jonathan Carter [[email protected]] [mail]
- Matthias Geiger [[email protected]] [mail]
- Sruthi Chandran [[email protected]] [mail]
- Santiago Ruano Rincón [[email protected]] [mail]
- Mattia Rizzolo [[email protected]] [mail]
- Vagrant Cascadian [[email protected]] [mail]
- Bas Wijnen [[email protected]] [mail]
- Michael Prokop [[email protected]] [mail]
- Guilhem Moulin [[email protected]] [mail]
- Ian Jackson [[email protected]] [mail]
- Andreas Tille [[email protected]] [mail]
- Didier Raboud [[email protected]] [mail]
Proposal H
Choice 8: Avoid the use of LLM: climate destruction is a deal breaker
LLM usage accelerates the destruction of our ecosystem (planet earth) and that is a deal-breaker
Using its power under Constitution section 4.1(5), the Project issues the following statement describing its current position regarding the use of LLM (Large Language Models) within Debian.
We don’t have a concern with the LLM technology as such: it’s just maths and enough training material, which outputs a very large set of numbers that we can use to infer new material from existing material. But we cannot talk about the LLM technology without talking about who is pushing for its usage, who is doing the training and how, what effect those have on our limited resources, and what are the economical and political motives of these.
Foremost we don’t understand how anyone can agree that global LLM usage accelerates the destruction of our ecosystem, planet earth, and not feel like that is a deal-breaker.
Too often the only response we’ve seen to this point was that other activities are also bad (some even worse) for the climate. How is this even an argument? The planet is burning, we should try to stop that as much and as fast as we can! If others aren’t stopping, isn’t that even more reason to compensate for their irresponsible actions?
And while we’re on that point: All of our users live on this planet, so protecting this planet is a matter of life and death for all of us. In other words, caring about something as big as climate change during our Debian work is self evident.
Debian recognizes that LLM also raise significant ethical, legal, technical, and social concerns. These include questions relating to copyright and licensing, privacy, the provenance and quality of generated material, the health of Free Software communities, and the long-term consequences of widespread reliance on LLM-generated content.
That said, we would like to emphasize that we condemn LLM (resource) usage but not LLM users. Fight the game, but not the players.
We also acknowledge that LLM usage can be hard if not impossible to detect and that Debian as a distribution cannot really impose LLM policies on other projects we package and distribute. Therefore this text is just a position statement. This said however, we appreciate the disclosure of LLM usage.
Accordingly, Debian encourages contributors to avoid the use of LLM where practical and to prefer human authorship, collaboration, and technical understanding over LLM-generated output.
At the same time, Debian has long relied on the judgment and responsibility of its contributors rather than prescribing individual workflows. Contributors remain responsible for everything they submit, irrespective of the tools used in preparing a contribution. Existing Debian standards regarding quality, correctness, licensing, and legal compliance continue to apply.
The Debian project has always recognized the commitment and professionalism of its members. All contributions are under the responsibility of the Debian Contributor making it, no matter the technology they have behind. We trust all Debian Developers, Maintainers and Contributors will continue to uphold the high quality values that have distinguished our project from its onset.
Please keep being excellent to each other and the only planet we have.
Data and Statistics
For this GR, like always, statistics will be gathered about ballots received and acknowledgements sent periodically during the voting period.
Majority Requirement
Proposal A needs a 3:1 majority, the other proposals need a simple majority.
Damn. I think it’s time for me to finally try Mint. Currently using 98% Ubuntu and 2% Windows because unfortunately there’s always some sort of Windows-only software I need for one reason or another. Yes I’ve tried Wine, I haven’t gotten it to work.
Do yourself a favor, do Linux Mint Debian Edition (LMDE). It’s got all the slickness of Mint, but the core is based on Debian instead of Ubuntu.
What Windows software do you find that isn’t working on Linux?
I can’t speak on behalf of him, but all my old CD-ROM games from the late 90’s and early 2000s don’t work on Linux, even with Wine. I’d love to get them working one day, so I’m watching this thread
Perhaps you could try bottles? I can’t speak on it’s usefulness or not, but something else to try. https://usebottles.com/
Thanks I’ll give it a go!
Primarily Solidworks, Fusion360, Altium, and Traktor Pro. Planning to start learning about video editing soon. Also games, though I’m planning to try Bazzite soon.
For me it’s Adobe. Stupid government and other forms I need for work. No available third party software will work on some of them, and I’ve tried them all. And for some dumb reason even having an Acrobat subscription won’t let me fill them out on their (extremely buggy, by the way) web app.
Was never able to get it working with wine.
That, and the compatibility between Libreoffice (or other Linux office software) and MS Word is still terrible. I’m quite sure that’s mostly due to shenanigans by MS. So if you have to swap documents back and forth between people (e.g. with track changes) it just doesn’t work as well. And, like Adobe, the web app version of Word seems to be kneecapped in a few key places.
Not him, but I almost never boot windows, but the other day I had to change a part of a a stupid powerpoint presentation that only could be done using Powerpoint desktop app. Wine has not done so well with it, Libreoffice wouldn’t do the thing right, and browser based office can’t touch ‘master slide’ content.
People love it for some reason but I cannot recommend. Newer laptops are trapped between no support under X and a half baked version under Wayland. And I didn’t find the community supportive enough.
May I recommend something fedora based? No dealing with Ubuntu or arch problems
I don’t see how a hat can fix that problem
Mint is based off of Ubuntu
There is a Debian based version called LMDE.
your saying snap-packages arent enough of a cluster-fuck to turn you off of Ubuntu?
Depending on the software, Winboat might be what you need.
What’s the software? Maybe a simple VM would do the job?
Sony DSLR apps are all non Linux. Some folks like me are forced to use Windows or Mac.
Solidworks, Fusion360, Altium, and a few others
I refuse to use Fusion360 because of what Autodesk did to Eagle (and my forever licence)
Autodesk are awful.
I hate it too. I use it to generate gcode for my CNC milling machine. Yes, Solidworks can also do that but it has a much steeper learning curve.
Pretty sure FreeCAD can also do that using the CAM workbench
Fusion360 actively attempts to block wine usage so you have to use a VM (or a single application VM like Winboat) for it to play ball. I haven’t had experience with any of the others though.
How does winboat work?
Makes a container windows install.
Do you have to see if the application you want to host is supported, or is it pretty compatible?
There are some windows softwares i want to run, but they don’t work in wine.
Its a full VM so the main downside is that GPU acceleration is 0 but you have full CPU speeds.
Also expect the VM to eat the typical 60gb + the applications of storage
Could we find Open Source alternatives? I’ve gotten good use of blender for these but I don’t know your use cases
I wish there were open source alternatives to Autocad, but freecad is doing its own thing and as someone who uses Autocad as part of her career it’s so much easier to keep a partition for when I need to throw something together than to learn how to make freecad work
Blender is world class, but the CAD softwares are leagues behind, unfortunately.
Whilst FreeCAD certainly doesn’t have feature parity with things like Creo, I firmly believe it is good enough for most people. I switched to only using that for all my contract work and personal projects and it’s generally been fine
That’s awesome! Do you have any videos you’d recommend for someone getting into it?
Very different use case. I make parts to 3D print and CNC mill. Lots of brackets and other mechanical parts.
Solid Edge is what had me keeping a Win10 instance for a while, but finally gave up on it after the OS kept corrupting itself. Started learning Onshape instead, which is web based and works under Linux. It’s a passable replacement for doing basic 3d sketches.
I’m curious how they will handle the Linux kernel itself, Linus has given the green light for AI slop.
Not sure why there are down votes, this is a legitimate concern. https://docs.kernel.org/process/coding-assistants.html
There’s a large amount of pro LLM users on .world who refuse any conversation that even slightly criticizes the technology. They don’t understand how it works while simultaneously using it to feed a false ego.
I can tell you I will get downvoted for this comment. I’ve already been banned from a few communities for arguing with these knuckle draggers. For the record I’m not even anti “AI”, just LLMs. Try explaining machine learning to them and watch their heads explode.
Yeah, that’s probably it
Do you in your heart of hearts believe this what you said here to be accurate?
Like I am not a fan of LLM slop coded projects, but I don’t think what you said can be honestly claimed to be accurate.
Ubuntu has a long and storied history of making bad decisions. Why should AI be any different?
There’s nothing about Ubuntu’s proposed use of AI that seems bad at all. Using it to improve accessibility is a good thing, especially when it’s open source, local-only, and sandboxed. I can’t see how having high quality speech-to-text would be a bad decision. Much of that stuff is probably not even considered “AI” by most people.
Arch users are forced to write their own AI?
AI - Archist Intelligence
(dude this is so cringe)
AI powered by the Arch Wiki and AUR
What are they going to do about upstreams that use it? Like the Linux kernel for example.
Oh, I didn’t know Ubuntu was a cuck OS. Good to know.
