Abstract
- Flathub bans AI-generated code for submissions and apps.
- The maintainer cites impolite, entitled submitters as the cause of the stern coverage.
- Current AI-tainted apps keep; new LLM-generated submissions can be rejected.
There is not any query about AI’s impact on coding. It has large ramifications on how other folks create apps and techniques, and curators of device have needed to imagine whether they permit AI-generated code of their initiatives. Linus Torvalds has already given the golf green gentle to AI-generated code (even if it nonetheless must be just right code), however a maintainer at Flathub has long past the other approach and banned it, as a result of other folks simply can not behave.
Flathub now not accepts AI-generated code
Other people simply were not being well mannered about it
Over on Mastodon, Flathub maintainer Bart Piotrowski introduced a brand new replace to Flathub’s LLM coverage. The brand new coverage “explicitly disallows AI utilization for each the submission procedure and packages being submitted.” Whilst Piotrowski had reservations about banning AI code, and he understands that coding will grow to be much less and not more human-driven as time is going on, other folks had been simply now not being very well mannered about it:
I consider it may be a useful gizmo in and outdoor FOSS; I was hoping we can see a bigger collection of apps the place authors made some effort past prompting an agent. In the meantime, the collection of unsightly interactions I have had with entitled submitters performing as though they had been bestowing their sensible device upon us idiots who’re rejecting it went during the roof within the closing month. I am drained.
The foundations aren’t being carried out retroactively, so if any person has already submitted an app that used AI-generated code, it will get to stick. Then again, other folks making an attempt to publish code made via an LLM don’t seem to be going to have a great time, particularly if they are impolite about it. Actually, it strikes a chord in my memory of how the protected trojan horse record channels for the Linux kernel free up applicants are being bombarded with trivial stories as a result of individuals are the usage of AI to seek out them and overstating their significance.

GNOME is cracking down on AI-generated code after updating its extensions tips
The reviewers have had sufficient.



