> Date: Wed, 29 Jul 2026 16:22:18 -0400 > From: Daniel Colascione <[email protected]> > > An LLM would be the perfect tool for making changelog commits, BTW. I *hate* > the things and over the years, I've kept a change local instead of bothering > to itemize in a commit message all the changes git already records that I've > done. > > Yet even this use of LLMs, which is just tedium , not skill, and which > (IANAL) has about as much copyright hook as a phone book (see Feist v Rural > Telephone) is prohibited. It's hard to interpret this prohibition as rooted > in practical considerations.
It is not prohibited. If you interpreted this incident as a prohibition, that is a misunderstanding. I apologize if my wording somehow caused that. The problem was not with the fact that the commit log message was generated with LLM assistance, the problem was in the Co-Authored-By heading as part of the log message, which could have been interpreted as meaning that the _code_ changes were produced with LLM assistance. We removed the original commit to avoid this confusion. The commit log messages have no legal significance (e.g., we modify the log messages as submitted by some contributors without changing the attribution, which does have legal significance). Thus, the (temporary) prohibition on using LLM-assisted contributions in the project does not include commit log messages, as it doesn't include any other text posted to this and other Emacs mailing lists. It only affects contributions of code and project documentation (the one that goes to manuals and doc strings). --- via emacs-tangents mailing list (https://lists.gnu.org/mailman/listinfo/emacs-tangents)
