On Thu, Oct 22, 2026 at 05:15:40PM -0500, Andrei Ionuţ Ştefan wrote:
>
> Setting that aside, the substantive question I'd still like an answer to,
> from anyone with more legal background who wants to take it: given that
> AFC-style analysis is apparently impractical to do credibly without
> litigation-grade work, is there a realistic middle ground for a project
> like this, or does "we can't practically rule this out" functionally
> disqualify it from archive inclusion as things currently stand? Daniel
> raised that distinction earlier in the thread and I'd genuinely like to
> understand where the real floor is — independent of which list, if any, is
> the right place to settle it.
I will note that if there is any human contribution, that part will be
copyrightable. under the license claimed by the human. So for
example, if I take some public-domain code, and modify it such that it
can be integrated into e2fsprogs, my changes to, say,
lib/ext2fs/crc32c.c are something that I can put under the LGPL. And
if someone takes a patch which was originally generated using LLM, and
then make changes to improve it, or when it modifies other code which
has human-licensed GPLv2 code, the resulting code is under the GPLv2.
The concern that people seem to have is a threhold question. If 99%
of the code is human-generated, and 1% LLM-generated or public domain
code, can we consider the entire code base under the GPL license?
What is if it's 60% / 40%? What if it's 99% LLM or public domain, and
1% human-generated GPLv2 changes?
I very much doubt this is something that debian-legal can answer....
but given that we do have projects that are partially LLM-generated
contributions, any policy that we come up with should ideally not
claim that say, the Linux kernel is not copyrightable because it
contains LLM-generated code, and thus must be dropped from Debian.
- Ted
