On Sat, 21 Mar 2026, Theo de Raadt wrote:

> I have looked at the diffs.
> 
> There is a claim that University of California holds copyright over large
> chunks of code which are new.  These are perhaps mostly copied, but have
> been changed in novel ways.  I didn't dig deep enough to decide if the
> changes are trivial or complicated, I just looked at the volume.
> 
> There is a different claim that you hold copyright over large chunks of
> new code.

This is IMO the essence of the problem here. This isn't using AI as a
code review, refactoring or merely mechanical tool, but instead using
it in context where it is writing code in (what appears to be) excess
of the originator's knowledge and skill. This code in question is highly
specific too, it's not like "go draw a stick figure of a person" and
more like "go paint Salvador Dali's Metamorphosis of Narcissus".

Who is the copyright holder in this case? It clearly draws heavily from
an existing work, and it's clear the human offering the patch didn't do
it. It's not the AI, because only persons can own copyright. Is it the
set of people whose work was represented in the training corpus? Was the
it the set of people who wrote ext4 and whose work was in the training
corpus? The company who own the AI who wrote the code? Someone else?

We don't know. The law hasn't caught up to the technology yet and we
can't take the risk that, when it does, it will go in a way that makes
use of AI-written code now expose us to legal risk. 

-d