GCC AI Policy Announcement

[Sam James]

Joseph Myers via Gcc <gcc@gcc.gnu.org> writes:

> On Wed, 29 Jul 2026, Richard Biener via Gcc wrote:
>
>> It does not cover the IMO much more severe issues of overwhelming 
>> reviewers and keeping GCC maintainable.
>
> Surely that's what the "All contributions must be submitted by a human who 
> understands the changes and is prepared to answer questions about them." 
> is about.

That's indeed the origin of the sentence. There were some concerns about
having stronger language and wasn't much consensus to go further than
what was in there.

There was also a feeling that:
1) some policy being published was necessary even if we refine it later,
   because the vacuum wasn't helpful for some cases (e.g. GSoC), and

2) addressing some issues would be better with examples and some
   practical experience with the policy as it stands today.

It's not going to surprise anyone that I lean towards the more
conservative side on these topics and would personally prefer a
stronger set of rules, but we needed to find common ground (and I
think we did a reasonable job at that).

There were also concerns about how enforceable a stronger policy would
be and I'd say some sense that the "not legally significant" threshold
was a useful proxy for now, in addition to being a prudent position to
take anyway.

Throughout, it was discussed that the policy would surely be revisited
if circumstances change (including demand from the community).

sam
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 418 bytes
Desc: not available
URL: <https://gcc.gnu.org/pipermail/gcc/attachments/20260729/d9ef3916/attachment.sig>