Re: Current policy on AI-generated code in NumPy

Matti Picus via NumPy-Discussion <[email protected]> Wed, 4 Mar 2026 17:24:27 +0200
Newsgroups gmane.comp.python.numeric.general
Message-ID <[email protected]>
--===============3853632816485937722==
Content-Language: en-US
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: 8bit

<!DOCTYPE html>
<html style="direction: ltr;">
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
    <style id="bidiui-paragraph-margins" type="text/css">body p { margin-bottom: 0cm; margin-top: 0pt; } </style>
  </head>
  <body bidimailui-charset-is-forced="true"
    bidimailui-detected-decoding-type="UTF-8" style="direction: ltr;">
    The NumPy steering council thanks the participants in this thread.
    We think it<br>
    has raised a several important points, which we propose to tackle<br>
    separately by opening a number of PRs to NumPy's PR template and<br>
    documentation (using, if needed, new, targeted threads on the
    numpy-discussion mailing<br>
    list).  Specifically,<br>
    <br>
    - A PR to the PR template to ask first-time contributors to
    introduce<br>
    themselves and how they use NumPy, as a way to signal that we value
    the<br>
    human connection.<br>
    <br>
    - A PR to the documentation which points contributors to guidelines<br>
    for responsible use of AI, such as the SciPy policy<br>
    <a href="https://scipy.github.io/devdocs/dev/conduct/ai_policy.html"
      rel="noreferrer" target="_blank"
data-saferedirecturl="https://www.google.com/url?q=https://scipy.github.io/devdocs/dev/conduct/ai_policy.html&amp;source=gmail&amp;ust=1772630470035000&amp;usg=AOvVaw2zasUsDMJlRYD9B4pK9wKd">https://scipy.github.io/devdoc<wbr>s/dev/conduct/ai_policy.html</a>.
    The PR<br>
    will mention the nice list Melissa put together<br>
    <a
href="https://github.com/melissawm/open-source-ai-contribution-policies?tab=readme-ov-file"
      rel="noreferrer" target="_blank"
data-saferedirecturl="https://www.google.com/url?q=https://github.com/melissawm/open-source-ai-contribution-policies?tab%3Dreadme-ov-file&amp;source=gmail&amp;ust=1772630470035000&amp;usg=AOvVaw3pTgpSso-wakte0zT4OZp1">https://github.com/melissawm/o<wbr>pen-source-ai-contribution-pol<wbr>icies?tab=readme-ov-file</a>.<br>
    <br>
    - Once we have a guideline, a PR to add checkboxes to the PR
    template<br>
    for disclosure in accordance with the guideline.<br>
    <br>
    - A PR to add another checkbox for taking responsibility for
    releasing<br>
    the code under NumPy’s license in accordance with the guidelines and
    other<br>
    documentation. (Done separately as the discussion so far suggests
    this may<br>
    need most thought in terms of phrasing and examples.)<br>
    <br>
    We envision that the PR template checkboxes and information will be
    used<br>
    as a tool for code reviewers to accept or reject the PR based on the<br>
    guidelines. The Steering Council remains available to help resolve
    any controversial disputes.<br>
    <br>
    We do not propose to reach a prescriptive yes/no on AI, since we
    feel like leaving the<br>
    actual decision in the hands of the reviewers makes it easier to
    evolve<br>
    policy with time.
  </body>
</html>

--===============3853632816485937722==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
NumPy-Discussion mailing list -- [email protected]
To unsubscribe send an email to [email protected]
https://mail.python.org/mailman3//lists/numpy-discussion.python.org
Member address: [email protected]

--===============3853632816485937722==--