After https://github.com/llvm/llvm-project/pull/172515, we have a new paragraph in LLVM policy about AI: > The one exception we reserve is for GitHub issues labelled with the “good first issue” label. These issues are selected by LLVM contributors to help newcomers get familiar with the code base. Thus, it makes no sense to fix them using AI tools. Using AI tools to fix issues labelled as “good first issues” is forbidden. We should add disclosure about it in the introduction note for developers to see clearly. --------- Co-authored-by: Reid Kleckner <rkleckner@nvidia.com>