Flathub moves to ban nearly all apps and submissions made with generative AI

Quoting: CatKiller

Applications or changes containing copyrighted, license-incompatible, or ethically questionable code are not allowed.

They're going to have to clean up the wording on this: all code is copyrighted.

in my opinion, they mean: everything that was stolen from GPL or proprietary software shouldnt be used .
permissive licences like mit, might be used.

"the reality is that you can expect less organically grown code as time goes on"

This.

To be honest all this is going to do is make people not be honest about using AI.. you people want transparency but the route your taking will only hide the fact people are using AI because 1000% bet you people can NOT tell human code from properly iterated and tested AI code, and now these coders because I again 1000% the people using AI properly are actual coders to be able to get the AI to produce code you won't be able to tell if its human or not

Last edited by tohur on 29 May 2026 at 5:11 pm UTC

I know it's an unpopular opinion on the Fediverse, but I do think LLMs are inevitable, and the reality is that you can expect less organically grown code as time goes on. I believe it can be a useful tool in and outside FOSS; I hoped we will see a larger number of apps where authors made some effort beyond prompting an agent. Meanwhile, the number of unpleasant interactions I've had with entitled submitters acting as if they were bestowing their brilliant software upon us idiots who are rejecting it went through the roof in the last month. I'm tired.

I think "LLMs are inevitable" is a terrible phrasing for a lot of reasons (it's like the "communism is great on paper, but..." thing - not about the stance, it's that people repeat this generic phrasing that doesn't actually make sense!), but I basically agree.

My current leaning is that it seems like LLM-based coding systems have a useful future. The problem is social/human. Corporations using it to gain power in extremely harmful ways...and individuals who have decided that simply using AI makes them superior to people with genuine skill/practice. They're more an extension of "AI artists" than they are "AI coders".

It is sensible and necessary to filter out AI submissions as a whole at this time.

I suppose this is also a good opportunity for Flatpak as a format. If an AI-based project distinguishes itself, people don't need to wait for Flathub to package/access it as a Flatpak, because they can be distributed outside of that centralized "app store". It raises a social barrier, but one that can be overcome.

Quoting: tohurTo be honest all this is going to do is make people not be honest about using AI.. you people want transparency but the route your taking will only hide the fact people are using AI because 1000% bet you people can NOT tell human code from properly iterated and tested AI code, and now these coders because I again 1000% the people using AI properly are actual coders to be able to get the AI to produce code you won't be able to tell if its human or not

you have defensive ai tools for that, like pangram and winston

Good. I like it when platforms take strong stances about AI.

It's his repo and his choices.
The contents of a repo are what makes or breaks it.
Some repos reject all closed source stuff, Apple rejects what it can't control, Many reject malware, github rejects nudifying AI, debian14 rejects non-reproducable packages
Many of these policies are perfectly reasonable and sometimes lawfully enforced, but they're all an integral part of the repo experience.

I think they're going to have to walk this back. It's far too restrictive; the way it's worded, if an LLM has ever even sneezed in the general direction of your code, your work is not allowed on Flathub. And for the great many developers who use LLMs to assist their coding without actually having those LLMs write the code (which is frankly the majority of developers now) it's simply not going to be tenable to ban their work outright.

I mean, maybe he wants to keep doubling down on this, he certainly can. But I think some major projects are going to start pulling their work in protest (KDE for example?)

I understand the sentiment and even mostly agree with it.

But the current way it is written is way too restrictive.
This is essentially going to kill Flathub if it remains like that.

Mostly because of this part:

Applications containing AI-generated or AI-assisted code, documentation, or other content are not allowed.

The vast majority of developers use at least AI-assistance in software development nowadays. A few years from now, you basically won't find anyone anymore not doing so (except a few luddites, I guess, but they'll find fewer and fewer gigs).

Besides being completely unable to verify that to begin with.

For example, if I use AI to answer me a bunch of questions about the code base in front of me, then make changes in the code on my own, that would still be AI-assisted.
Same thing if you want to switch some file to a different coding guideline, but let AI do it. Result would be identical to what anyone would do manually, but suddenly it isn't kosher anymore.
Or if you changed 100 lines of code, AI detected a typo and fixed a line for you.
Or even if you ran it on some code to detect issues and fixed those yourself.
The amounts of legitimate use cases for AI, without letting AI do the entire coding work and produce garbage, is huge.

Scratch the AI-assisted part of that requirement and I think it would be quite reasonable.

And this part, too, is just strongly eyebrow-raising:

These submissions can be rejected without any further review.

So someone as much as believes something was even just AI-assisted, bam! - that's it. 🤣

Quoting: Arthur PrazeresYou have defensive ai tools for that, like pangram and winston

They can of course detect some blatantly obvious stuff like entire commits & pull requests, etc. being made with AI.

But you cannot detect assistance, as it may not even lead to any code being written by AI at all.

Last edited by TheSHEEEP on 30 May 2026 at 10:39 am UTC

In the end, the issue is of course the one of filtering a deluge of mostly or purely AI-generated slop from legitimate commits.

The amount to sift through on public repositories is huge, and that is generally done by a human - who would then spend significant amounts of time looking at garbage code "produced" by some vibers.
Understandable that that's not a job anyone would want to do willingly, many probably not even for money.

The best workaround I see is to use the defensive tools available, as mentioned previously.
Those can at least filter out the most obvious offenders, and maybe raise flags for some other stuff so humans can then look at it.
It'll still be more than it used to be prior to AI tools...

But yeah, ultimately this issue has not been resolved in any satisfactory way for projects that anyone can send pull requests to.

Maybe the only solution is to no longer allow pull requests from any rando and only allow actually vetted team members to send PRs.
And then maybe collect potential changes "from outside" elsewhere, also in a way that should only allow actual humans to suggest them there.

But how?
🤷‍♂️

Last edited by TheSHEEEP on 30 May 2026 at 10:33 am UTC

Quoting: TheSHEEEPThe amounts of legitimate use cases for AI, without letting AI do the entire coding work and produce garbage, is huge.

Is it really "legitimate" to burn the world for such simple tasks? We are in the heaviest phase of the climate change. We cross the 1.5K difference right now and instead of consuming less energy as we should, we do the opposite with so called "AI" and burn more resources than ever, just to do what? Any 0.1K more destroys the earth in a way that is not repairable anymore, even if we could lower the temperature by 0.1K 10 years later. I know the fight is lost, people start using LLM for any tasks in their lives, they also drive cars and fly airplanes as there would be no issue at all. But everyone who does not participate and tries to reduce the footprint should be applauded. So I do with Flathub, even if their reasons are different, the result is the same.
Climate change is the worst part of LLM - even worse than all the real slop productions everyone hates or the LLM tools build in everywhere (yes I speak about you, Copilot). Even worse than the increased prices (and I know many people are just not able to buy replacements - I feel with you, my PC is also not allowed to break any time soon).

Sure, it will probably not visible to anyone, but that is not the point. I think we all would benefit if people would stop using it for any single task and only use it in the few tasks where it really makes a huge difference. I am even using DuckDuckGo lite to avoid LLMs getting activated and burning resources in first place. I am also using LLMs, but as least as possible (and most of the time, because slop-webpages making classic research impossible from time to time).

And for those who are still not convinced (which tells a lot of those people): do you really want to bind yourself to LLMs, which once day become locked behind paywalls that are impossible to pay with an average income? Right now we have 20$/€ per months subscriptions. In future you pay every single token on models that can cost into the thousands or even tenths of thousands per months depending on the usage. I know a journalist who created costs of multiple hundreds of Euro per day (without using them in parallel). So you should at least ask yourself how much you really want to bind yourself to these tools for the moment you have to pay the real prices, plus the profit margin of the company that wins this LLM-race.

Quoting: PlayingOnLinuxphone

Quoting: TheSHEEEPThe amounts of legitimate use cases for AI, without letting AI do the entire coding work and produce garbage, is huge.

Is it really "legitimate" to burn the world for such simple tasks? We are in the heaviest phase of the climate change. We cross the 1.5K difference right now and instead of consuming less energy as we should, we do the opposite with so called "AI" and burn more resources than ever, just to do what? Any 0.1K more destroys the earth in a way that is not repairable anymore, even if we could lower the temperature by 0.1K 10 years later. I know the fight is lost, people start using LLM for any tasks in their lives, they also drive cars and fly airplanes as there would be no issue at all. But everyone who does not participate and tries to reduce the footprint should be applauded. So I do with Flathub, even if their reasons are different, the result is the same.
Climate change is the worst part of LLM - even worse than all the real slop productions everyone hates or the LLM tools build in everywhere (yes I speak about you, Copilot).

You do you. 🤷‍♂️

Quoting: PlayingOnLinuxphonedo you really want to bind yourself to LLMs, which once day become locked behind paywalls that are impossible to pay with an average income? Right now we have 20$/€ per months subscriptions. In future you pay every single token on models that can cost into the thousands or even tenths of thousands per months depending on the usage.

People who do bind themselves to AI usage because they lack the skills to do stuff without it, will indeed have a hard time once/if that comes to pass.
But that is a self-solving issue - those people are just gonna drop out then.
The rest will use a useful tool as far as it will remain useful to them, and affordable. Once it isn't those things anymore, they'll just continue without it.

Although I'm not sure if that will happen first or the bubble burst...

Quoting: TheSHEEEPYou do you. 🤷‍♂️

In other words: "I do not take response for my actions and I am fine with being part of the people destroying the planet and killing 95% of all species on earth." 🤦‍♀️ Great to live on costs of those who are not even born... (I did not even say to stop using LLM entirely).

People who do bind themselves to AI usage because they lack the skills to do stuff without it, [...]

I did not even say "those who lack the skill", I say those who don't want to lose the benefit of LLMs, but will not be able to pay it any longer, once they have to pay the real costs.

@Playingonlinuxphone you can set ddg search assist to "on demand", it does exactly what it implies (no nagging). Also is there a local alternative to search assist or similar tools, say just to re-rank the top 25 results? They are sometimes miraculous.

Quoting: PlayingOnLinuxphone

Quoting: TheSHEEEPYou do you. 🤷‍♂️

In other words: "I do not take response for my actions and I am fine with being part of the people destroying the planet and killing 95% of all species on earth." 🤦‍♀️ Great to live on costs of those who are not even born... (I did not even say to stop using LLM entirely).

Yeah, you'll have to excuse me while I do not follow in this particular apocalypse prophecy cult after the last 200 or so also did not come to pass.
It's kind of gotten a bit boring, honestly, the constant fear mongering, moralistic finger wiggling and the proselytizing.

You seem to have accepted already that people are no longer listening, so I'm not sure what exactly you are trying to achieve.
Maybe this is fun to you, in which case: You do you.

Quoting: PlayingOnLinuxphoneI did not even say "those who lack the skill", I say those who don't want to lose the benefit of LLMs, but will not be able to pay it any longer, once they have to pay the real costs.

Does it matter?
The outcome is the same - it'll either go away or (more likely) become used a lot less if the costs exceed the use.
Parts of the AI-related issues will definitely resolve themselves.
Others, not so much.

Last edited by TheSHEEEP on 30 May 2026 at 7:09 pm UTC

Quoting: TheSHEEEPYeah, you'll have to excuse me while I do not follow in this particular apocalypse prophecy cult after the last 200 or so also did not come to pass.

I never did myself, but I always was very aligned with natural science. And what's happening is as obvious as not to touch high power electricity and anyone saying the opposite is just an excuse to continue destroying the environment. The data is public, everyone can study it just like free software, because it is free knowledge. Not trusting in science and not checking itself just makes you sound like these people who believe in apocalypse prophecies.

What I am saying is no prophecy. I speak about real data based estimates. They can be less bad or worse ... the data of last 50 years were always too conservative (the real world data became worse). That also shows that we can compare old prediction with real world changes and know how well our year 2100 prediction is working. I usually would not care if it only affects yourself (just like these apocalypse prophets), but we even see (or experience) negative consequences right now.

It's kind of gotten a bit boring, honestly, the constant fear mongering, moralistic finger wiggling and the proselytizing.

It has nothing to do with fear. It is just a reality and I could write you a book and deliver sources from scientists to each single argument I would write. Boring is this "but my car, my AI, my other thing that I really really want to use, no matter what harm I do with it ... I mean I do no harm at all"...

You seem to have accepted already that people are no longer listening, so I'm not sure what exactly you are trying to achieve.
Maybe this is fun to you, in which case: You do you.

I accepted it, yes. But not because a few people do not listening. The main issue is the capitalism. We would not even talk about climate change any more, if we had a less destructive economy system worldwide. All climate change related fake news come from people who want to make money or for those stupid enough to trust them - so called [useful idiots](https://en.wikipedia.org/wiki/Useful_idiot).

Why I am still trying to avoid the worst case? Pretty easy: for the principle.
1) I don't want to be guilty and want to say to next generations "sorry, I tried my best, it was just not enough".
2) Who give up already has lost. When I play PvP on games and it is "clear" that my side loses the battle, 1/10 games will still be won, because I never give up. If all people of my team continue trying, we still win every third "lost game". So if I continue to try and enough people continue trying too, the chance to survive as species is not too bad. The damage until this date still persist for ever (species that are dead, stay dead). Also all the other damage that will recover will require thousands of years (even the damage we already did with 1.5K as the increase amount of nature catastrophes per year).
3) Every 0.1K more also makes our own life worse. Foot becomes more expensive, conflicts in the world can increase (climate refugees etc) and things like nature catastrophes will also increase in amount and intense, which also affects my home land. Enough good reasons to continue the fight against stupidity and egoism.

Anyway, that is a bit too off-topic here. The point is: we do not need to avoid LLM everywhere, but the increase in LLM usage this year will exceed 100 Million tons CO2 emission (probably far above) while all green energy that is used for these data centers is not available for other households (so they even green-wash this number). If we would already produce 100% green energy everywhere, LLM would not be such a big issue as is. But we did not even replace oil with electricity..

Does it matter?

Tell you me. I don't rely on LLM right now and my little usage can be done with a local model that costs my own power bill. People who heavily use it to code stuff (which requires the more 'til most potent models) will probably run in more issue if they build their whole workflow around it.