From 3ae898a37b6b1c25188a2f2b547c05a39ee8abd8 Mon Sep 17 00:00:00 2001 From: Michael Rothgang Date: Wed, 6 May 2026 22:59:13 +0200 Subject: [PATCH 1/5] feat(policy): tweak AI policy Lift the blanket ban on new contributors using AI --- but emphasize that low-quality PRs can be closed without discussion. This allows flexibility to accept useful PRs, but retains enough leeway to deal with slop quickly. --- templates/contribute/index.md | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/templates/contribute/index.md b/templates/contribute/index.md index 6aa5d24f9..6641a05f7 100644 --- a/templates/contribute/index.md +++ b/templates/contribute/index.md @@ -67,7 +67,10 @@ Note that using an LLM when writing comments on github or Zulip is not allowed: Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use and if it is truly necessary. -We do not accept pull requests opened by new contributors where code is LLM-generated. Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of April 2026, AI-written code fails to meet that bar by a large margin. Getting code to mathlib's standards requires understanding and writing Lean code by hand. If you just want to help and not put in the learning effort, making a PR to mathlib is counterproductive: the effort required from the mathlib maintainers is larger than the benefit, because the time used to improve the quality of the code will not result in a better quality in future PRs. More generally the reviewer team may close a PR containing low-quality code if it appears to be AI-generated. If we notice that you open several PRs without putting in this learning effort and without adhering to our community ethical standards, we retain the right of banning you both from opening new PR's and from the Zulip chat. +Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team reserve the right to summarily close without comment any low quality PR, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. +If we notice that you open several PRs without putting in this learning effort and without adhering to our community ethical standards, we retain the right of banning you both from opening new PR's and from the Zulip chat. + +Getting code to mathlib's standards requires understanding and writing Lean code by hand. If you just want to help and not put in the learning effort, making a PR to mathlib is counterproductive: the effort required from the mathlib maintainers is larger than the benefit, because the time used to improve the quality of the code will not result in a better quality in future PRs. If you use artificial intelligence (such as, by using github's copilot mode, asking an LLM like [ChatGPT](https://chatgpt.com/) or using an agent like [Codex](https://openai.com/codex/), [Claude](https://claude.ai/), [Gemini](https://gemini.google.com/app), or even Lean-dedicated agents like [Aristotle](https://aristotle.harmonic.fun/)), please explain this in the PR description. Explain which tool(s) you used and how you used it. This provides useful context for reviewers: tools make different mistakes than humans, so knowing this makes it easier to spot common errors. From ee106f7ba24163355c7be21bcdc89ac1ad3e426c Mon Sep 17 00:00:00 2001 From: Michael Rothgang Date: Thu, 7 May 2026 15:24:10 +0200 Subject: [PATCH 2/5] Update templates/contribute/index.md --- templates/contribute/index.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/templates/contribute/index.md b/templates/contribute/index.md index 6641a05f7..c9c29ce04 100644 --- a/templates/contribute/index.md +++ b/templates/contribute/index.md @@ -65,7 +65,7 @@ change on Zulip and, when significant consensus among reviewers is reached, to o Note that using an LLM when writing comments on github or Zulip is not allowed: use your own words. -Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use and if it is truly necessary. +Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use. When reviewing PRs, we are particularly concerned that the pedagogical value of the reviewers work is wasted if there is not a human contributor actively learning. Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team reserve the right to summarily close without comment any low quality PR, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. If we notice that you open several PRs without putting in this learning effort and without adhering to our community ethical standards, we retain the right of banning you both from opening new PR's and from the Zulip chat. From 1a266f0c19ed80c944f231147c2f352ca825dbc5 Mon Sep 17 00:00:00 2001 From: Michael Rothgang Date: Fri, 8 May 2026 08:49:15 +0200 Subject: [PATCH 3/5] Update templates/contribute/index.md --- templates/contribute/index.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/templates/contribute/index.md b/templates/contribute/index.md index c9c29ce04..5e5305fd4 100644 --- a/templates/contribute/index.md +++ b/templates/contribute/index.md @@ -67,7 +67,7 @@ Note that using an LLM when writing comments on github or Zulip is not allowed: Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use. When reviewing PRs, we are particularly concerned that the pedagogical value of the reviewers work is wasted if there is not a human contributor actively learning. -Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team reserve the right to summarily close without comment any low quality PR, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. +Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team will summarily close without comment any low quality PR produced using LLMs, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. If we notice that you open several PRs without putting in this learning effort and without adhering to our community ethical standards, we retain the right of banning you both from opening new PR's and from the Zulip chat. Getting code to mathlib's standards requires understanding and writing Lean code by hand. If you just want to help and not put in the learning effort, making a PR to mathlib is counterproductive: the effort required from the mathlib maintainers is larger than the benefit, because the time used to improve the quality of the code will not result in a better quality in future PRs. From 6896f8413bb312c434d409bcdd2d8a42e8c8a2d0 Mon Sep 17 00:00:00 2001 From: Michael Rothgang Date: Fri, 8 May 2026 08:49:26 +0200 Subject: [PATCH 4/5] Update templates/contribute/index.md --- templates/contribute/index.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/templates/contribute/index.md b/templates/contribute/index.md index 5e5305fd4..e0ffd5681 100644 --- a/templates/contribute/index.md +++ b/templates/contribute/index.md @@ -68,7 +68,7 @@ Note that using an LLM when writing comments on github or Zulip is not allowed: Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use. When reviewing PRs, we are particularly concerned that the pedagogical value of the reviewers work is wasted if there is not a human contributor actively learning. Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team will summarily close without comment any low quality PR produced using LLMs, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. -If we notice that you open several PRs without putting in this learning effort and without adhering to our community ethical standards, we retain the right of banning you both from opening new PR's and from the Zulip chat. +If we notice that you open several PRs without putting in this learning effort or without adhering to our community ethical standards, we will suspend (or permanently ban) you both from opening new PR's and from the Zulip chat. Getting code to mathlib's standards requires understanding and writing Lean code by hand. If you just want to help and not put in the learning effort, making a PR to mathlib is counterproductive: the effort required from the mathlib maintainers is larger than the benefit, because the time used to improve the quality of the code will not result in a better quality in future PRs. From e96d9821f9212217dba46b13219fe9fe18a45cd8 Mon Sep 17 00:00:00 2001 From: Michael Rothgang Date: Sat, 9 May 2026 08:52:51 +0200 Subject: [PATCH 5/5] Final wording tweak --- templates/contribute/index.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/templates/contribute/index.md b/templates/contribute/index.md index e0ffd5681..75af6cedf 100644 --- a/templates/contribute/index.md +++ b/templates/contribute/index.md @@ -67,7 +67,7 @@ Note that using an LLM when writing comments on github or Zulip is not allowed: Using artificial intelligence tools to generate code is becoming more and more common. While this can be practical, their use also poses ethical, ecological, legal and social concerns. We recognise that there are strong differences in opinion on this topic. That said, while individual action alone will not address these concerns, we ask you to consider the effects of your AI use. When reviewing PRs, we are particularly concerned that the pedagogical value of the reviewers work is wasted if there is not a human contributor actively learning. -Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, AI-written code fails to meet that bar by a large margin. Members of the review team will summarily close without comment any low quality PR produced using LLMs, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. +Mathlib intentionally has very high standards (on generality, integration with the remaining library and maintainability, including code style). As of mid-2026, code written by an AI without the supervision of a Lean subject expert fails to meet that bar by a large margin. Members of the review team will summarily close without comment any low quality PR produced using LLMs, especially if the author has made little effort to directly engage in the community in a discussion about its merits before opening the PR. If we notice that you open several PRs without putting in this learning effort or without adhering to our community ethical standards, we will suspend (or permanently ban) you both from opening new PR's and from the Zulip chat. Getting code to mathlib's standards requires understanding and writing Lean code by hand. If you just want to help and not put in the learning effort, making a PR to mathlib is counterproductive: the effort required from the mathlib maintainers is larger than the benefit, because the time used to improve the quality of the code will not result in a better quality in future PRs.