GenAI policy

[Mte90]

Our project is growing and talking with @Ph0enixKM we need to address this thing.

We are using GenAI, as maintainers but also our contributors AI to produce code so we need a sort of policy in PR that specify if the code is by human or not, what AI it was used and the model version.

I was thinking that we can improve the PR template when someone opens a new one with this details.

I am against a 0 GenAI policy not because I am using too (and I am not so much Rust skilled, as I usually worked in other things in the project) but because that somehow is the future and we need to address instead of rejecting.

We are using AI tools like Copilot and CodeRabbit for code review and they helped a lot the development and also the review.

[Ph0enixKM]

I'm think that we should only accept AI code that was modified by a human or at least fully and closely reviewed

[callframe]

I'm in favor of this:

systemd repository on Github docs/CONTRIBUTING.md:

We expect everyone contributing to systemd to fully own their contribution, be able to reason about it, be able to explain why things were done a particular way and act as the full owner of that code. AI tools are treated the same as traditional tooling like sed, awk or coccinelle.

For the purpose of this project, AI tools CANNOT be treated as author, co-author or be credited in any way that would suggest any ownership over the contribution.

The contributor should have done all the thinking, planning and understanding of the changes needed to resolve an issue or implement a new feature prior to using automated tooling to perform the grunt work.

Unguided use of those tools or the inability to prove understanding of the code contributed will result in a loss of trust in that contributor by project maintainers which can then lead to exclusion from any further contribution to the project.

As with any other submissions, authors are responsible for doing due diligence and ensuring their submissions are compatible with the project's license as documented in LICENSES/README.md.

[Mte90]

So if a user creates a PR with co-ownership from claude or copilot we need to asks to remove it.

We already had this kind of things.

[callframe]

In my opinion, if the majority or all of the changes have been done by AI then code quality will degrade further also without human supervision bugs, that could be found, go unnoticed, memory usage increases and performance degrades

[Mte90]

Probably also make sense to have an AGENTS.md file that asks to the AI if it is opening a PR to disclose the model and the version.

[Tirito6626]

i also like the systemd approach, i dont think that using AI for the compiler is a good idea; but using AI for amber code itself isn't much of an issue yet and can sometimes be useful for creating example code/small parts for the main code, since the language isn't that complex yet