I'm not a BSD dev. Actually, I kind of suck at writing code. But I do like to write other material.

There's a list of projects and their stances on AI slopcode that someone keeps posting-
https://codeberg.org/ethical-foss/open-slopware#bsd-based

The definition of "permissive" contains this gem:
Opening the door for others to use gen AI counts as a "Permissive AI Policy".

So if some human from outside the project finds a bug using an LLM, submits the bug after manually reviewing and confirming, the project is just supposed to ignore the bug, just so they can pass a purity test? Because that is what I'm getting from this requirement.

All I'm seeing is credit provided as submitted.

"But but but, then that means they support <insert batman slapping some sense meme here>...". No, it doesn't mean that. If I wrote a book, and someone found errors in the book, and I confirmed those errors existed, I'm going to correct them. That is reasonable. I may find it personally distasteful that someone used an LLM in doing so, and I can mention that when providing credit for the editorial change, but that doesn't mean I sought out LLM assistance or even wanted it.

I will not just leave that glaring error there. I won't write the fix in some other language, like Klingon, just to say I didn't accept that knowledge of the error. I'm also not going to go through some weird convolutions to then add a bunch of overhead on top of the editorial review, just to then magically reproduce a structural change in my copyrighted work that now imagines me somehow coming up with exactly the same fix in a way that allows me to deny ever have received that report.

There isn't any cognitive dissonance involved in my also saying I don't use LLMs or AI, and I don't write with them, or create anything using them. Nothing about that is "permissive", nor does it "condone" using AI or LLMs.