Codeberg Bans LLM Training and Vibe-Coded Projects to Protect FLOSS
Protecting our FLOSS commons from LLMs
Our members voted to strictly prohibit using Codeberg data for training Large Language Models and to ban 'vibe-coded' projects. We believe LLMs drain our infrastructure, inflate hardware costs, and erode the trust essential for free software collaboration. By rejecting these technologies, we aim to preserve a sustainable, human-centric ecosystem where genuine community effort thrives over automated, resource-hungry generation.
Using LLMs to work with your code gives you a kick of adrenaline, but you are often alone, working with a statistical machine that turns energy into code.
- tptacek
This makes sense. The largest project hosted on Codeberg today is, what, Zig? Zig is also openly hostile to LLM agents. It's a a clear identity for the site: a home for projects that identify as anti-LLM.
I disagree with basically all of the logic they're using here, but that's the point: not every site needs to be for every person. Clearly there are quite a few software developers who agree vehemently with the stance they're taking. They should have a community hub, and now they do.
- croemer
Since the blog post doesn't actually include the change to the TOS and only links to it at the very end, here it is:
> 7. You must not share projects that mostly consist of code written by "generative AI"-tools (including services such as Claude, OpenAI Codex). Such projects having an unclear copyright status (see requirements § 2 (1) 1 and § 2 (1) 3) and furthermore have little safeguards to ensure that they do not include harmful code (c.f. § 2 (1) 5).
https://codeberg.org/Codeberg/org/commit/96fac426a32d1ba91ff...
- arjie
Seems fine. A git host is not a particularly hard piece of software to self-maintain and there’s always GitHub for most things.
I can understand why. The cost to the host is entirely in volume of code, and creating volume is cheap now. Without some constraining measure they will have lots of AI code. Even just by itself that would make it impossible to have.
It’s also a cultural thing. This is for that niche of developers. That’s all right. Communities have their own (often-hyperspecific) rules. HN certainly does and we’re all here.
- arkwin
Developers who hate LLMs: You are not creating software according to the exact rules I have made up in my head about how software should be made, so I'm going to build a wall and hate you and the code you make.
That's what the "protecting the community from LLMs" sounds like.
There is plenty of perfectly acceptable software created using LLMs. There isn't a good way to detect LLM usage in the creation of software. It's like saying "don't cheat" on a test and then calling out anything that looks suspicious? Too many comments explaining the code? AI. Lots of if statements? AI. I think it looks too much like AI? Must be AI.
I don't know, seems like a losing battle to me. I'm glad people voted on it. And you can be against AI for boiling the ocean, but plenty of people are using open offline models to code.
As for the copyright argument, I doubt it will come up in the average person's life. It really hasn't even been argued well in court either.
Finally, what is "mostly consist of code written by generative AI tools"? Ok, let's think that out. The Linux kernel? Probably not a lot of AI-generated code. The Linux kernel in 5-10 years? Probably the majority of code written by AI. Or any piece of software over time at this rate.
My 2 cents.
- somnium_sn
It is fine to have corners of the internet that run in a particular way.
My issue however, is that this is changing the rules on people. Codeberg initially was a FLOSS alternative to big company Github. Now it's more narrow, an alternative to Github and only for people who don't want to use AI. What's the next step? This is the problem, you either a platform and be, within legal reasons, agnostic, or you are trying to be a specific community. It is unclear to me if the specific community that Codeberg wants to be today will be the one it wants to be in a year from now, because clearly it's changing.
I cannot rely on a service provider that does change that quickly.
- TheChaplain
I don't really see a problem with this decision (or the one regarding crypto), they were voted on by the active members of Codeberg, simply a democratic outcome.
- evantbyrne
From their recent announcements it seems Codeberg is operating as a club that just so happens to have a git UI. Perhaps that is amusing if you're playing along, but I can't imagine why I should want to actually host anything at an org with this kind of governance.
- tomxor
There's an interesting tension between "censorship" and "ideology" here.
On the one hand I think organisations taking pure ideological stands is healthy for society as a general behaviour, rather than appealing to the masses and business insentives as GitHub would; on the other hand that can technically manifest as censorship in some cases.
But that's ok as long as we have more than one of a type. The answer to disagreement with one small organisation or projects ideology is to make another one, which is exactly what the author has done in the extreme (host my own damn git). This is the same thing that democratizes FOSS, projects can be ideological dictatorships, that's ok because they can be forked and no one is forced to stay. It also allows individual projects to efficiently follow their ideology or principles without the machinery of democracy getting in their way... It's the best of both.