{"thread":{"id":"66208","subject":"[PATCH] gitk: discourage AI contributions","startedAt":"2026-08-23T10:04:24Z","lastAt":"2026-08-26T21:55:51Z","messageCount":4,"participants":["Johannes Sixt","Weijie Yuan","brian m. carlson"],"isPatch":true,"patchVersion":1,"patchTotal":null},"messages":[{"id":"551059","messageId":"b4ce8b09-5faa-4e6e-af38-5384a71ae699@kdbg.org","threadId":"66208","inReplyTo":null,"subject":"[PATCH] gitk: discourage AI contributions","fromName":"Johannes Sixt","fromEmail":"j6t@kdbg.org","sentAt":"2026-08-23T09:25:42Z","receivedAt":"2026-08-23T10:04:24Z","isPatch":true,"body":"Since the provenance of code generated by AI cannot be tracked with\nconfidence, which would be necessary to sign off a contribution, tell\ncontributors not to use AI.\n\nSigned-off-by: Johannes Sixt <j6t@kdbg.org>\n---\n Following a recent brief discussion[1] it looks like I should not\n accept code for Gitk that was generated by AI even if the submitter\n understands all of it. Should I discourage AI contributions outright\n like this or would it overshoot the mark?\n\n -- Hannes\n\n [1] https://lore.kernel.org/git/aooRdiVdjovWSFiG@fruit.crustytoothpaste.net/T/#md44580584f30c9b3e3c9e23250baf08b7fd7fb0c\n\n README.md | 7 ++++---\n 1 file changed, 4 insertions(+), 3 deletions(-)\n\ndiff --git a/README.md b/README.md\nindex 2e30746..c402788 100644\n--- a/README.md\n+++ b/README.md\n@@ -79,13 +79,14 @@ are also accepted.\n All commits must be signed off (use `git commit --signoff`) and should\n have commit messages prefixed with `gitk:`.\n \n-Email Patches\n--------------\n-\n Send patches to git@vger.kernel.org and CC j6t@kdbg.org. See the Git\n project's [patch submission guidelines](https://git-scm.com/docs/SubmittingPatches)\n for detailed instructions on creating and sending patches.\n \n+Please do not submit work that originated from AI, because this project\n+is down-streamed to the Git project, which has a rather\n+[strict AI guideline](https://git-scm.com/docs/SubmittingPatches#ai).\n+\n License\n =======\n \n-- \n2.55.0.704.g7397dd7565\n\n"},{"id":"551070","messageId":"aorxVo_6_U1ceaKm@wyuan.org","threadId":"66208","inReplyTo":"b4ce8b09-5faa-4e6e-af38-5384a71ae699@kdbg.org","subject":"Re: [PATCH] gitk: discourage AI contributions","fromName":"Weijie Yuan","fromEmail":"wy@wyuan.org","sentAt":"2026-08-23T13:10:46Z","receivedAt":"2026-08-23T13:10:55Z","isPatch":true,"body":"[+cc Johannes Schindelin]\n\nOn Sun, Aug 23, 2026 at 11:25:42AM +0200, Johannes Sixt wrote:\n> Since the provenance of code generated by AI cannot be tracked with\n> confidence, which would be necessary to sign off a contribution, tell\n> contributors not to use AI.\n> \n> Signed-off-by: Johannes Sixt <j6t@kdbg.org>\n> ---\n>  Following a recent brief discussion[1] it looks like I should not\n>  accept code for Gitk that was generated by AI even if the submitter\n>  understands all of it. Should I discourage AI contributions outright\n>  like this or would it overshoot the mark?\n> \n>  -- Hannes\n> \n>  [1] https://lore.kernel.org/git/aooRdiVdjovWSFiG@fruit.crustytoothpaste.net/T/#md44580584f30c9b3e3c9e23250baf08b7fd7fb0c\n\nI followed the discussion between Brian and Junio, and I have no\nintention of further expanding the discussion about the legal and\ntechnical issues surrounding the use of AI/LLM. However, I would like to\npoint out that several of dscho's recent commits seem to carry an\nAssisted-by trailer, and, unless I missed something (possibly), nobody\nappears to have raised the issue or strongly objected to it. Of course,\nI believe dscho also fully understood the code he submitted.\n\n(This is getting a little away from the DCO legal issue that Junio and\nBrian were discussing.)\n\nSo how do we distinguish between \"Generated\" and \"Assisted\"? Perhaps the\nLLM merely assisted him with some investigative work rather than\nactually writing any code.\n\nMeanwhile, in the Linux kernel community, probably the most prominent\nuser of the DCO, within it, Linus also expressed his views [1] on the\nuse of LLMs a few weeks ago, while the Linux Foundation has published\nguidance on the legal issues surrounding the use of AI. [2]\n\nAlthough I know that the Git and Linux kernel communities do not follow\nexactly the same rules, Jonathan Corbet updated the kernel's\nrequirements for disclosing LLM use a couple of days ago, namely:\n\nAssisted-by: LLM [TOOL1] [TOOL2]\n\nPerhaps this gives us a way to tell whether a patch contains content\ngenerated by an LLM. (But if we want to ban it entirely, please\ndisregard this part.)\n\nThanks,\nWeijie\n\nEmm, I´m not sure which thread would be the right place to continue this\ndiscussion, sorry.\n\n[1] https://lore.kernel.org/all/CAHk-=wi4zC+Ze8e+p3tMv8TtG_80KzsZ1syL9anBtmEh5Z40vg@mail.gmail.com/\n[2] https://www.linuxfoundation.org/legal/generative-ai\n"},{"id":"551261","messageId":"ao6N0O2wmPgXwM9n@wyuan.org","threadId":"66208","inReplyTo":"aorxVo_6_U1ceaKm@wyuan.org","subject":"Re: [PATCH] gitk: discourage AI contributions","fromName":"Weijie Yuan","fromEmail":"wy@wyuan.org","sentAt":"2026-08-26T06:55:12Z","receivedAt":"2026-08-26T06:55:24Z","isPatch":true,"body":"On Sun, Aug 23, 2026 at 09:10:54PM +0800, Weijie Yuan wrote:\n> [+cc Johannes Schindelin]\n\nPlease directly ignore my previous reply. Sorry for the noise.\n\nThanks.\n"},{"id":"551319","messageId":"ao9g3AaGJGQR0N_p@fruit.crustytoothpaste.net","threadId":"66208","inReplyTo":"b4ce8b09-5faa-4e6e-af38-5384a71ae699@kdbg.org","subject":"Re: [PATCH] gitk: discourage AI contributions","fromName":"brian m. carlson","fromEmail":"sandals@crustytoothpaste.net","sentAt":"2026-08-26T21:55:41Z","receivedAt":"2026-08-26T21:55:51Z","isPatch":true,"body":"On 2026-08-23 at 09:25:42, Johannes Sixt wrote:\n> +Please do not submit work that originated from AI, because this project\n> +is down-streamed to the Git project, which has a rather\n> +[strict AI guideline](https://git-scm.com/docs/SubmittingPatches#ai).\n> +\n\nI think this seems like the right thing to do.\n-- \nbrian m. carlson (they/them)\nToronto, Ontario, CA\n"}]}