On Wed, Jul 8, 2026 at 11:46 AM Andreas Enge <[email protected]> wrote: > > Am Wed, Jul 08, 2026 at 01:43:08AM +0200 schrieb Denis 'GNUtoo' Carikli: > > So can we do something like that and wait (and not accept code/data > > made with LLMs) until GNU releases its rules? > > I would answer this question with "no". If this GCD does not pass, then > the status quo prevails, and this status quo does not state that we do > "not accept code/data made with LLMs". So it would be at the discretion > of the individual committer.
The GCD process does not preclude the project from temporarily rejecting LLM-authored legally significant contributions while waiting on legal guidance from the GNU Foundation. This is a perfectly reasonable step to take while we improve our tracking of copyright status. This would need to be added to the user documentation and pull request template. > This is exactly the topic of the "cost of reverting" section in the GCD: > "What would be costly to revert is the *lack* of any form of regulation > on genAI use in Guix." A GCD is significantly, one might say here infinitely, harder to revert than a simple change to the documentation. There has been no effort to build consensus. > We can always add more contributions written by AI in the future, but it > would be close to impossible to remove them. Agreed on the former but GNU has stated that the latter is not necessary.
