KDE's proposed LLM policy: use AI freely but don't disclose it, mocked as Fight Club rules

BLUECOW009 · x · 2026-09-21

KDE has published a draft LLM usage policy (GitLab issue #187) built around a "Don't be lazy" golden rule: AI use is encouraged as long as there's a real human in the loop, and obviously lazy LLM-generated contributions may be ignored or closed.

The most controversial rule is non-disclosure: contributors are told not to add "Assisted-by: [some LLM]" commit tags (the opposite of the Linux kernel's practice), on the grounds that proper usage should produce output indistinguishable from what you could produce yourself, and the tag is just free advertising for the vendor. Critics likened it to Fight Club: "the first rule of developing KDE with AI is you do not talk about developing KDE with AI."

Other points: no throwing throwaway LLM changes at maintainers as "drafts", no vibe-coded changes you couldn't make yourself, no using AI disclosure to excuse low-quality work, and text generation with LLMs is generally discouraged.

Original post →

More from coding & agent

coding & agent channel →