AI policy

[SashankBhamidi]

• shared AI policy as a webpage under pycal.org (easier to make updates and to be followed across projects)

[stevepiercy]

• Use https://icalendar.readthedocs.io/en/latest/contribute/index.html#artificial-intelligence-policy as a starting point.
• Compare against https://nlnet.nl/foundation/policies/generativeAI/, and merge relevant bits into icalendar's.
• Any other references? @tobixen you mentioned you have an AI policy somewhere in chat?

[stevepiercy]

I reviewed the documents. Here are the additions I think we need under https://icalendar.readthedocs.io/en/latest/contribute/index.html#responsible-ai-use.

If you use AI in your work:

- You must disclose the AI's terms of use.
- You must ensure that outputs are not reconstructed from copyrighted sources.
- You are expected to understand and be able to explain design and code decisions.
- You must distinguish git commits that add generated code and include the used model and prompts in your commit message.
- You must disclose that you used AI in your change log entry.

I'll create a PR in icalendar for all of us to review, and ask for NLnet to look at it, since it is critical to receive payment from NLnet.

[stevepiercy]

See related PR:

collective/icalendar#1312

After that PR is approved, then we can move its content over to pycal.org and provide a link to it. One fewer thing to maintain in icalendar.

[tobixen]

My last edition of my AI-POLICY is at https://github.com/python-caldav/caldav/blob/v3.2-preparations/AI-POLICY.md

[niccokunzmann]

I am sharing general thoughts:

icalendar has an AI policy. I am happy to use that for my projects.

So, we should probably check out the following:

• Can we use icalendar's AI policy on the website?
• Does the policy need adapting to be used here?
• Should we just link to it?
• Should projects link to it individually?

Also:

• For the AI policy, it might be important to name the requirements that go into it, list them. E.g. NLNet, icalendar, first-timers, maintainer protection
• icalendar might have a different standard than other projects. Just copying it over here might be a bit a problem. See Update AI policy, per NLnet document. collective/icalendar#1312 (comment)
• What do other projects do? If we feature a good AI Policy, maybe others also have a stance on that and discussions, what AI is good for and what it is not, how to use it responsibly.

[stevepiercy]

If we want to receive funding from NLnet without the risk of termination of the grant, then we absolutely should use icalendar's AI policy for all projects that receive its funding.

I'd prefer to maintain one policy in one place, with pycal.org as the authoritative source, and all projects that receive NLnet funding link to it. No duplication please, as that adds to the maintenance burden.

I want to keep this as simple as possible. The more detail, the less likely it will be read, be understandable, or be followed, especially when translated. Remember, the substantive factors that go into an AI policy are the following.

1. Limited community members' volunteer time
2. Funding requirements
3. Legal matters, including copyright

[SashankBhamidi]

My idea is that we will have an AI policy probably at pycal.org/ai-policy, and across every project, we will link to this policy instead of RTD AI policy that is doc specific, easy to maintain.

What does everyone feel about that?

[stevepiercy]

My idea is that we will have an AI policy probably at pycal.org/ai-policy, and across every project, we will link to this policy instead of RTD AI policy that is doc specific, easy to maintain.

What does everyone feel about that?

How is that different from what I suggested?

[SashankBhamidi]

My idea is that we will have an AI policy probably at pycal.org/ai-policy, and across every project, we will link to this policy instead of RTD AI policy that is doc specific, easy to maintain.
What does everyone feel about that?

How is that different from what I suggested?

No, I meant I agree with you. That's what I had in my mind when I created this issue.

[niccokunzmann]

It might be worth putting the policy into its own repository. It can be then included in the main page as an updatable submodule. Revisions can be clearly tagged without the problem of the website. Discussions are kept focussed. Since AI is evolving rapidly, I would like to be able to provide a link and say: We follow PyCal-AI-PolicyV1 and with that allow people to consciously update it. Could be ai.pycal.org/v1.

What I would want to avoid is a clash between flexibility to update and projects & people to understand the upgrade. Clear diff is also possible.

[stevepiercy]

Put text under the title.

# AI policy

v1.0.0 2026-04-09

Look at the page git history.

Simple, effective, nothing more to add.

[tobixen]

https://nlnet.nl/foundation/policies/generativeAI/archive/feedback/ is informative