April 13, 2026 at 6:08 AM, "Amos Jeffries" <[email protected] 
mailto:[email protected]?to=%22Amos%20Jeffries%22%20%3Csquid3%40treenet.co.nz%3E
 > wrote:



> 
> On 12/04/2026 08:07, jbranso wrote:
> 
> > 
> > April 11, 2026 at 5:38 AM, "Amos Jeffries" wrote:
> > 
> > > 
> > > On 11/04/2026 01:31, jbranso wrote:
> > > 
> >  I'm assuming that the Hurd's current policy is not to accept AI 
> > contributed code.
> >  Opperating under that assumption, I'll add a patch to the wiki documented
> >  that we are ok with AI to debug the Hurd codebase, but NOT to write code.
> >  Thanks,
> >  Joshua
> > 
> > > 
> > > I recall there being issues with peoples patches to some of the Hurd 
> > > related GNU projects not being merged due to copyright assignment issues.
> > > 
> >  I don't believe that this is an issue. I have been contributed to the hurd 
> > wiki for years (without copyright assignment).
> >  When I submitted my first contribution to the Hurd manual, Samuel would 
> > not merge my patches until I had assigned copyright.
> > 
> Exactly. You were able to assign **your** copyrights.
> 
> IFF the AI is "a legal person/entity", then they/it have to come and assign 
> any permission for Hurd to use their code. Per normal policy the patch is not 
> permitted, yet.
> 
> If the AI is "just a tool", then things become a question of who has the 
> copyrights for the generated patch.
> 
> I am not sure if the foundational question of whether AI is generating new 
> content, or copying others work has been settled in legal circles. Let alone 
> whether there is a Hurd policy or view about it.
> 
> > 
> > So I'm pretty sure Samuel would catch anyone trying to contribute without 
> > assignment.
> > 
> That is not the point.
> 
> The point is about what to publish in the wiki about **why** AI patches are 
> rejected.

If we come to a consensus on what I should write on the wiki, then let me know!
 
> > 
> > > 
> > > I'm not sure who is in charge of Hurd legal matters, but it would be 
> > > worth checking with them for anything like the above, to document as 
> > > relevant for AI generated code.
> > > 
> > >  HTH
> > >  Amos
> > > 
> >  I do wonder what the FSF will encourage GNU package maintainers about A.I. 
> > I assume that they do not want AI contributions, but I
> >  do not know if they have released an official statement or not.
> >  I know Linus just allowed AI contribution to Linux: 
> > https://docs.kernel.org/process/coding-assistants.html
> >  Thanks,
> >  Joshua
> > 
> Thanks for that link. That kernel policy has some interesting implications 
> and foresight in the wording.
> 
> HTH
> Amos
>

Reply via email to