Re: [Biopython] Generative AI policy for contributions to Biopython

--===============8338974931288567676==
Content-Type: multipart/alternative; boundary="000000000000bf516506503da643"

--000000000000bf516506503da643
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Hi Peter,

I lean towards limiting but not banning AI contributions. Fully automatic
contributions by agents should be banned, but the reality is that we cannot
police what people use to work on code. We can, however, ban contributions
from people that violate our "rules".

Beyond a policy, some other ideas:
- review our contribution guidelines, to make it explicit to incoming devs
what we expect of them (maybe this is redundant with the policy);
- maybe include an AGENTS.md file (https://github.com/agentsmd/agents.md)
to at least provide some guidance to people wanting to use agents to help
them work on our codebase;l and flag when they do so (tag, etc);
- and add an explicit checkbox/agreement to the pull request template
indicating that the human user is responsible for the contribution under
penalty of ban from future ones.

Cheers,

Jo=C3=A3o



A sexta, 24/04/2026, 12:16, Peter Cock <[email protected]> escreveu=
:

> On Fri, Apr 24, 2026 at 12:51=E2=80=AFPM Markus Piotrowski
> <[email protected]> wrote:
> >
> > Dear Peter,
> >
> > While I also tend to strongly restrict or even forbid the usage of AI i=
n
> > Biopython,
>
> Thank you - and as one of top top contributors that carries weight
> and is encouaraging.
>
> > ... I wonder how you really can prevent this. A careful submitter
> > can mask the AI signs in his/her code so that I will go undetected. So
> > wouldn't it be better to allow the usage under strong restrictions and
> > conditions (and I agree with the those that you have mentioned,
> > including the "good first issues" in your other e-mail) to encourage
> > potential contributors to be transparent about the of use of AI?
> >
> > I'm unsure about this, but I wanted to include this topic into the
> > discussion.
> >
> > Best
> > Markus
>
> That is a real worry, and I think Andrew and Bastian have given good
> answers.
>
> Thank you,
>
> Peter
> _______________________________________________
> Biopython mailing list  -  [email protected]
> https://mailman.open-bio.org/mailman/listinfo/biopython
>

--000000000000bf516506503da643
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<div dir=3D"auto">Hi Peter,<div dir=3D"auto"><br></div><div dir=3D"auto">I =
lean towards limiting but not banning AI contributions. Fully automatic con=
tributions by agents should be banned, but the reality is that we cannot po=
lice what people use to work on code. We can, however, ban contributions fr=
om people that violate our &quot;rules&quot;.</div><div dir=3D"auto"><br></=
div><div dir=3D"auto">Beyond a policy, some other ideas:<br>- review our co=
ntribution guidelines, to make it explicit to incoming devs what we expect =
of them (maybe this is redundant with the policy);</div><div dir=3D"auto">-=
 maybe include an AGENTS.md file (<a href=3D"https://github.com/agentsmd/ag=
ents.md">https://github.com/agentsmd/agents.md</a>) to at least provide som=
e guidance to people wanting to use agents to help them work on our codebas=
e;l and flag when they do so (tag, etc);</div><div dir=3D"auto">- and add a=
n explicit checkbox/agreement to the pull request template indicating that =
the human user is responsible for the contribution under penalty of ban fro=
m future ones.</div><div dir=3D"auto"><br></div><div dir=3D"auto">Cheers,</=
div><div dir=3D"auto"><br></div><div dir=3D"auto">Jo=C3=A3o=C2=A0</div><div=
 dir=3D"auto"><br></div><div dir=3D"auto"><br></div></div><br><div class=3D=
"gmail_quote gmail_quote_container"><div dir=3D"ltr" class=3D"gmail_attr">A=
 sexta, 24/04/2026, 12:16, Peter Cock &lt;<a href=3D"mailto:p.j.a.cock@goog=
lemail.com">[email protected]</a>&gt; escreveu:<br></div><blockquot=
e class=3D"gmail_quote" style=3D"margin:0px 0px 0px 0.8ex;border-left:1px s=
olid rgb(204,204,204);padding-left:1ex">On Fri, Apr 24, 2026 at 12:51=E2=80=
=AFPM Markus Piotrowski<br>
&lt;<a href=3D"mailto:[email protected]" target=3D"_blan=
k" rel=3D"noreferrer">[email protected]</a>&gt; wrote:<b=
r>
&gt;<br>
&gt; Dear Peter,<br>
&gt;<br>
&gt; While I also tend to strongly restrict or even forbid the usage of AI =
in<br>
&gt; Biopython,<br>
<br>
Thank you - and as one of top top contributors that carries weight<br>
and is encouaraging.<br>
<br>
&gt; ... I wonder how you really can prevent this. A careful submitter<br>
&gt; can mask the AI signs in his/her code so that I will go undetected. So=
<br>
&gt; wouldn&#39;t it be better to allow the usage under strong restrictions=
 and<br>
&gt; conditions (and I agree with the those that you have mentioned,<br>
&gt; including the &quot;good first issues&quot; in your other e-mail) to e=
ncourage<br>
&gt; potential contributors to be transparent about the of use of AI?<br>
&gt;<br>
&gt; I&#39;m unsure about this, but I wanted to include this topic into the=
<br>
&gt; discussion.<br>
&gt;<br>
&gt; Best<br>
&gt; Markus<br>
<br>
That is a real worry, and I think Andrew and Bastian have given good answer=
s.<br>
<br>
Thank you,<br>
<br>
Peter<br>
_______________________________________________<br>
Biopython mailing list=C2=A0 -=C2=A0 <a href=3D"mailto:Biopython@biopython.=
org" target=3D"_blank" rel=3D"noreferrer">[email protected]</a><br>
<a href=3D"https://mailman.open-bio.org/mailman/listinfo/biopython" rel=3D"=
noreferrer noreferrer" target=3D"_blank">https://mailman.open-bio.org/mailm=
an/listinfo/biopython</a><br>
</blockquote></div>

--000000000000bf516506503da643--

--===============8338974931288567676==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Biopython mailing list  -  [email protected]
https://mailman.open-bio.org/mailman/listinfo/biopython

--===============8338974931288567676==--