I’d like to raise a process question regarding the application of the AI policy to reviewed pull requests.
A recent PR <https://github.com/sympy/sympy/pull/29051> addressing a long-standing performance issue in degree() was closed with the label “AI slop,” despite the following: - The change was technically reviewed (by myself) - The correctness and performance improvements were verified - The author disclosed AI usage and stated that finding the source of the problem and design were manual - The PR appears to comply with current AI policy as written (or at least I don't see the violation) I’m not arguing that the change must be merged. My concern is procedural: when a pull request has received substantive technical review and endorsement, it seems problematic for it to be summarily dismissed without a technical rationale. If there is no review and the code clearly doesn't address the problem in a meaningful way then I don't see a problem with closing it after a couple of days. As the AI policy currently stands, it permits AI-assisted contributions provided the author understands and takes responsibility for the code. And having reviewed the code, I can't see why labelling it as “AI slop” is a sufficient basis for closure in the absence of technical objections. I’d like to ask whether we should clarify policy or process here, for example by distinguishing between unreviewed submissions and those that have received substantive technical review. At minimum, it would be helpful to document whether reviewed PRs are expected to receive a technical disposition, even when AI assistance is involved. I’m happy to help draft a clarification if that would be useful. /c On Wednesday, January 28, 2026 at 7:39:54 AM UTC-6 [email protected] wrote: > Dear all, > > Thanks for all the input. The AI policy is now live: > > https://docs.sympy.org/dev/contributing/ai-generated-code-policy.html > > Jason > moorepants.info > +01 530-601-9791 <(530)%20601-9791> > > > On Fri, Jan 16, 2026 at 8:45 AM <[email protected]> wrote: > >> I like it. If some reviewer rejects an AI generated PR / issue he/she has >> a “policy reason” to point to. >> >> B.t.w.: I did not once see that a submitter, whose PR was rejected on AI >> grounds, objected to this – apparently it *was* AI generated. >> >> >> >> Peter >> >> >> >> *From:* [email protected] <[email protected]> *On Behalf Of *Jason >> Moore >> *Sent:* Friday, January 16, 2026 8:22 AM >> *To:* [email protected] >> *Subject:* Re: [sympy] AI generated pull requests >> >> >> >> Dear all, >> >> >> >> We now have a second draft of the AI policy: >> >> >> >> https://github.com/sympy/sympy/pull/28941 >> >> >> >> based on a round of feedback. Please review the 2nd draft and leave any >> more feedback. >> >> >> >> I would like to have it merged after this coming round of feedback. >> Please have a look. >> >> >> >> Thanks, >> >> >> >> Jason >> >> moorepants.info >> +01 530-601-9791 <(530)%20601-9791> >> >> >> >> >> >> On Wed, Jan 14, 2026 at 1:46 PM Jason Moore <[email protected]> wrote: >> >> Dear all, >> >> >> >> I opened this PR with a draft of a SymPy AI code contribution policy: >> >> >> >> https://github.com/sympy/sympy/pull/28941 >> >> >> >> I suggest that people review and I will improve the text based on the >> feedback there until we have some consensus. >> >> >> >> This will not solve all issues associated with the onslaught of AI slop, >> but can at least get some kind of policy codified that we can use to help >> fend off the slop. >> >> >> >> Jason >> >> moorepants.info >> +01 530-601-9791 <(530)%20601-9791> >> >> >> >> >> >> On Thu, Jan 1, 2026 at 2:29 PM Sangyub Lee <[email protected]> wrote: >> >> > Let's do a simple test. Instead of commenting these PRs by typing text >> in, let's just attach an image containing the comment. This is no problem >> for human beings, but I expect AI-bots to fail in understanding the >> comment, unless they are connected with an OCR or use vision-language >> models. >> >> I don't think that it is difficult problem to solve than using Captcha >> for submitting a PR. That does not 100% solve the problem but at least >> raises barrier to professional spammers and scammers who can bypass that. >> Many social media services should have endured much more chaos than SymPy >> or Github for this stuff, so we can use their experiences. There are always >> more stricter options like policies allowing stricter face/ID verified >> accounts. >> Github may not have submitting captcha before PR, but we can implement >> something like CI check with 3rd party captcha service for new contributors. >> >> On Wednesday, December 31, 2025 at 12:31:10 AM UTC+1 Oscar wrote: >> >> This one here looks like a real AI bot: >> https://github.com/sympy/sympy/pull/28862 >> >> -- >> You received this message because you are subscribed to the Google Groups >> "sympy" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected]. >> To view this discussion visit >> https://groups.google.com/d/msgid/sympy/5810a4bc-b237-4ee7-a707-cb29cddea2d2n%40googlegroups.com >> >> <https://groups.google.com/d/msgid/sympy/5810a4bc-b237-4ee7-a707-cb29cddea2d2n%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> >> -- >> You received this message because you are subscribed to the Google Groups >> "sympy" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected]. >> To view this discussion visit >> https://groups.google.com/d/msgid/sympy/CAP7f1AiQqbW5CL%3DOyk6h3jnBCjrN4oBbv_wW4kxHt-yy7U3zPg%40mail.gmail.com >> >> <https://groups.google.com/d/msgid/sympy/CAP7f1AiQqbW5CL%3DOyk6h3jnBCjrN4oBbv_wW4kxHt-yy7U3zPg%40mail.gmail.com?utm_medium=email&utm_source=footer> >> . >> >> -- >> You received this message because you are subscribed to the Google Groups >> "sympy" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected]. >> > To view this discussion visit >> https://groups.google.com/d/msgid/sympy/061b01dc86bc%240d11db80%2427359280%24%40gmail.com >> >> <https://groups.google.com/d/msgid/sympy/061b01dc86bc%240d11db80%2427359280%24%40gmail.com?utm_medium=email&utm_source=footer> >> . >> > -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/sympy/2dcb1858-4cef-4016-9eb3-025f5969d4b9n%40googlegroups.com.
