Create AIPOLICY.rst - #5388

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 86.43%. Comparing base (23aa995) to head (99e941f).
⚠️ Report is 14 commits behind head on master.

@@            Coverage Diff             @@
##           master    #5388      +/-   ##
==========================================
+ Coverage   85.56%   86.43%   +0.87%     
==========================================
  Files         282      282              
  Lines       59616    59616              
==========================================
+ Hits        51011    51532     +521     
+ Misses       8605     8084     -521     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

Thank you Iddo - discussions on #5241 should benefit from a concrete draft of a more permissive policy.

This will need to intergrate with our CONTRIBUTING.rst file (at a minimum cross-referencing each other, perhaps even staying with one file?).

I'd also suggest something like #5334 adding an AI declaration check-box to our pull request template, which can reference this new file (assuming you keep it separate from CONTRIBUTING.rst).

What about the "good first issue" topic? I would think that experienced contributers will be more responsible in their usage of AI in their PRs, but "good first issues" are especially labelled for beginners and maybe we should insist on no-AI here?

As for "good first issue" -- the strict AI ban on GFI appears already in CONTRIBUTING.rst. @peterjc has suggested placing AIPOLICY.rst within CONTRIBUTING.rst, and if we do that then this issue is moot.

If it's about AI, it should be included in the AI policy. It wouldn't even harm if it's duplicated under Contributing and AI policy. There is a good chance that people who want to submit an AI written PR will carefully read the AI policy and skip the Contributing part.
This is also a good candidate for (additionally) mentioning in an "Allowed, discouraged and forbidden use of AI" chapter.

Thank you Iddo - discussions on #5241 should benefit from a concrete draft of a more permissive policy.

This will need to intergrate with our CONTRIBUTING.rst file (at a minimum cross-referencing each other, perhaps even staying with one file?).

Putting the AI policy in CONTRIBUTING.rst sounds logical. AI use policy is a subset of the contribution policy, after all.

I'd also suggest something like #5334 adding an AI declaration check-box to our pull request template, which can reference this new file (assuming you keep it separate from CONTRIBUTING.rst).

Suggestion:
"I have read the AI policy in [FILE] and agree to abide by it. I confirm that this pull request is performed by a human and not by autonomous software such as an AI agent"

I confirm that this pull request is performed by a human and not by autonomous software such as an AI agent"

The explicit mention of autonomous software and AI agent may be misunderstood that only this case is discouraged but e.g. copy & pasting an AI generated PR message is OK.

"I confirm that this pull request (communication) is written by a human." ?

I posted on the mailing list comparing the CoMaps/Forejo policy (which BioPerl is looking at) with NumPy/SciPy (both based on SymPy): https://mailman.open-bio.org/pipermail/biopython/2026-October/017166.html

Both stress human communications and no autonomous agents. The NumPy/SciPy text is more explicit about copyright, but neither really emphasises scholarship and attribution. That will come up, particularly with non-trivial changes (it has occasionally in the pre-AI era where a fix was at some level guided by another implementation). Since the AI tools don't track their training data for attribution, one might argue a human in the loop has to try to check providence - and that's hard, but I think any AI-permissive policy ought to articulate who is expected to do that too.

Should we make our PR template leaner in tandem with this? Otherwise we have 4 checkboxes?

I'd propose something like a mix of numpy's PR template and ours:

Closes #...

# PR summary
# First time contributor introduction
# AI Disclosure

[ ] I hereby agree to dual licence this and any previous contributions under both the Biopython License Agreement AND the BSD 3-Clause License.
[ ] I have read the CONTRIBUTING.rst file, have run pre-commit locally, and understand that continuous integration checks will be used to confirm the Biopython unit tests and style checks pass with these changes.
[ ] I have added my name to the alphabetical contributors listings in the files NEWS.rst and CONTRIB.rst as part of this pull request, am listed already, or do not wish to be listed. (This acknowledgement is optional.)

We might also revisit the existing checkboxes, e.g. the second one could be shortened to "I have read the CONTRIBUTING guidelines". We could add the AI policy linked there. And the third checkbox there too, leaving this only about licensing.

Yes, we now have four checkboxes as of #5334 (adding an AI declaration). Any more would be a burden (having seen this on several other projects - even ticking them all with [x] gets tedious). That will need tweakinng as part of agreeing an AI policy.