Lobsters

omg! ubuntu!

XDA Developers

Flathub, the main Flatpak app store for Linux, has updated its policy to ban AI-generated or AI-assisted code in new app submissions. The ban also covers AI-generated documentation and automated submissions.
The decision came from maintainer Bart Piotrowski, who was candid about what actually pushed him over the edge: it wasn’t the code quality, it was the submitters. A surge of entitled developers pushing vibe-coded apps through GitHub PRs, then getting dismissive or rude when reviewers pushed back, made the volunteer review queue genuinely miserable to manage. Piotrowski described the volume of low-effort submissions as “depressing.”
Piotrowski acknowledged some personal ambivalence about the ban. He sees LLMs as tools that aren’t going away, and he’s not philosophically opposed to them. But the behavior patterns that came with the AI-generated submissions left him little choice.
A few practical details worth noting:
<ul>
<li>Apps already on Flathub with AI-generated code are grandfathered in and won’t be removed</li>
<li>Exceptions may be made for mature, well-maintained projects that can demonstrate genuine commitment - things like community engagement, a consistent release cadence, and working CI systems</li>
<li>The ban covers both the apps themselves and the submission process</li>
</ul>
Flathub isn’t alone here. GNOME has made similar moves, and there’s been ongoing frustration in the Linux kernel community over AI-generated bug reports that waste maintainer time.
The core problem isn’t really about AI - it’s that volunteer-run review queues have a finite capacity, and a flood of low-effort submissions burns through that capacity fast. When the submitters are also rude about it, the calculus becomes pretty simple.

Collections

Flathub has updated its submission policy to ban AI-generated or AI-assisted code, documentation, and automated submissions for new apps. The decision was driven primarily by a surge of low-effort, vibe-coded submissions paired with rude behavior from submitters toward volunteer reviewers. Maintainer Bart Piotrowski acknowledged personal ambivalence about LLMs but cited the unsustainable burden on the volunteer review queue. Existing apps with AI-generated code are grandfathered in, and exceptions may be made for mature, well-maintained projects. GNOME has made similar moves, and the Linux kernel community has faced comparable frustration with AI-generated bug reports.

Flathub bans LLM-generated code in new app submissions, citing rude submitters and low-effort slop

I’m curious to know how they detect AI generated code/documentation

At the same time I agree with the decision and absolutely hate it.

That would make anyone frustrated.

Totally fair. The ban isn’t AI itself, but against the rude behavior and spam clogging up the system. Protecting the volunteers making this work is the right call

I wonder if they set up an LLM/Agent to triage the PRs and automatically reject AI Slop based on some criteria, the Vibe Coders would thing “Cool, the AI rejected me, so its ok”