Proposal: Code of Conduct Policy on AI‑Generated CSAM #10349
Replies: 5 comments 14 replies
|
Create a PR as a starter - #10365 |
|
No codes of conduct. Ever. Get lost. |
I'm pretty sure a code of conduct is not on the roadmap, even for something nobel like this. |
|
I know this PR is in bad faith (see https://x.com/SootySec/status/2096350564229550413?s=20 / https://archive.is/BF28G) but I don't think anybody would oppose some kind of rule about forbidding this in the project since it really is not illegal everywhere in the world. It just shouldn't be a code of conduct. |
|
I see... |
|
Happy for it not to be in a code of conduct but in a different set of rules. "A rose by any other name would smell as sweet" |
|
Why did you delete your tweet I linked? Here's another link I guess --> https://archive.is/BF28G |
|
not deleted anything, although it has disappear and I have a label message in twitter now which is weird. Glad you managed to save a copy as I didn't save it. |
|
most of the other message from other people have also disappeared my notification are full of loads of "This post was deleted by the post author" |
|
AH yes because a linux distro needs rules about AI content that has nothing to do with the distro itself. You should be contacting every model creator, and provider no a linux distro. |
|
I want to start with this thread, because the answers I got when I asked for priors and a motive were not a causal argument: https://x.com/tepri_x/status/2096377077792670088 The claim being made is: P1. DHH wants Elon / X to use Omarchy. C does not follow. P1 is about who DHH would like as a user of a distro. P2 is about abuse on a giant hosted platform. Neither shows that Omarchy is a venue for this material, that Omarchy users are doing it, or that a markdown rule here would stop it. What was never supplied:
That is a category error. X is a hosted network with accounts, uploads, and a ToS it can enforce. Grok is a hosted model with bans and servers. Omarchy is a local Linux install. Microsoft can write ToS because it controls identity, updates, and licensing. None of that maps onto an MIT-licensed desktop. Two problems with trying to enact the rule anyway: Definition. CSAM is not treated the same around the world. Many places distinguish material involving real people from virtual / fictional content; Japan is the obvious example. “AI-generated CSAM” has not been defined here. You cannot agree to, much less enforce, a rule whose object has not been specified. Enforcement. Banning people from official channels is reasonable. Maintainers already have that power, CoC or not. Claiming the project can grant or revoke “permission” to use the OS is not. This is FOSS. Anyone can clone, fork, or install it without an account and without agreeing to a file in this repo. A use-ban cannot be audited or revoked on machines the project does not control. It only binds people already visible in official spaces, which means it functions as a political test, not as prevention. If the actual request is “don’t put this in official assets or community spaces, and ban people who try,” say that. That is already enforceable. It does not require a CoC, a world-use ban, or the Elon argument. Frankly, I do not think we'd need a rule for that either, in the repo or in the community, as it is common sense that it would be followed. This is nothing more than CoC-theater, and I think it is being proposed for purposes not in good-faith. I would like to remind anyone reading this that these same slim rules slipping into communities where someone like this gets a presence has often slid into unwelcome territory. It is a fallacy to assume it will happen here too, but given the patterns of the past, particularly within the FOSS community, I think it would be best to heed doing so with caution and stick with common sense. That being said, I would not myself have a problem if not having assets like this in the project or community was a rule, to me, it doesn't feel like it would matter either way. |
There are multiple "live" issues which include people simple asking X to do X in creative way but also the fact X uses previous CSAM to generate now CSAM content which under Japan law would be illegal. https://www.theguardian.com/technology/2026/sep/03/elon-musk-ai-grok-child-porn-lawsuit |
|
Do you have evidence that the current maintainers do not already share that conscience, and would not remove this kind of content from the repo or official channels? That is common sense. Obviously no project wants that material in its tree or community spaces. What they cannot do is control who runs the OS. That part has to come out. A use-ban is incompatible with an MIT distro and is unenforceable. “X uses previous CSAM to generate new CSAM” is an empirical claim about another product’s pipeline. State the evidence or drop it. You do not get to treat it as a given and then skip from that to “Omarchy needs a use-ban.” |
|
Also that article is a complaint, not a verdict. Valid evidence of a lawsuit, but not of the fact you stated. |
So Grok giving you answers about it's master intentions are gospel but the evidence of a child sexual abuse victim should be treated with suspicion even when backed up by verifiable evidence. 
|
|
There is a difference between a newly filed complaint with attorneys saying C3P matched hashes of images on X to a known series, and that a court actually found that xAI trained on CSAM or that “X uses previous CSAM to generate new CSAM” as settled fact. It's a good reason to take the allegation seriously, but it's not the latter, calling that "suspicion of the victim" is a dodge. Even if every allegation in that suit were later proved, you would still have shown a problem at Grok/X, not that Omarchy is a venue for it. |
Uh oh!
There was an error while loading. Please reload this page.
I believe Omarchy should adopt a clear Code of Conduct addressing the prohibition of AI‑generated child sexual abuse material (CSAM). The following points outline what the policy should cover:
AI‑generated CSAM must not be included in the project in any form, including wallpapers, themes, assets, or any other contributions.
Omarchy should not permit the use of its operating system by individuals or organisations involved in creating, distributing, or facilitating AI‑generated CSAM.
This should be a straightforward, non‑controversial position within the technology community. Given recent discussions around Omarchy Kids and the broader risks associated with user‑generated content, it is important that we formally define and enforce these boundaries.
Clear guidance is necessary because global laws and regulations around AI‑generated CSAM are currently inconsistent, and in some jurisdictions insufficient. A well‑defined policy will help protect the project, its contributors, and its users.
All reactions