How to use AI to contribute

Canonical’s Open Documentation Academy (CODA) is a place for people to discover open source through documentation contributions. This guide explains how we think about Artificial Intelligence (AI) in the context of CODA contributions and how you can use AI tools as a contributor.

Why it matters

AI tools – Large Language Models (LLMs), code generators, grammar assistants – are now part of many people’s everyday life. We don’t ask you to put them aside when you contribute to CODA, but we do ask you to think carefully about how you use them.

CODA exists because open source should be an opportunity to learn and collaborate with others. When you contribute to a CODA issue, the main priority should be your learning and growth, not simply completing the contribution itself. The aim is to build skills, gain confidence, and become part of a community. AI can support that process, but it should not replace the experience of learning through contributions.

We want your contributions to be something you are proud of. Not because they are perfect but because they are yours.

How to use AI in CODA

We ask every contributor who chooses to use AI to follow three principles.

Honesty

If you use AI assistance when making a contribution, disclose that fact in your pull request description. Tell us which AI tool you used and how you used it.

For example: I used [tool name] to help check the grammar and organize the structure of this tutorial. All technical content and wording are my own. This transparency helps your mentor understand how to support you.

Learning

CODA issues are created for people to learn from. If you cannot complete an issue without an AI tool doing the most of the work, do not try to force it with AI. You are unlikely to learn much from the process, and you may not be able to properly verify the result. Instead, ask your mentor for support, or choose a different issue.

Responsibility

You must understand and stand behind what you submit. AI can help you produce better work, but the work must remain yours to own. Even if you used AI tools, you must ensure that your contribution still reflects your judgement, skills, and competence.

What we do not accept

AI usage in CODA should be done responsibly, as a tool that supports your work and growth, not as a replacement for thinking, crafting, and tinkering. There are ways of using AI that undermine the purpose of CODA. Some examples that would be rejected by the mentor include:

• Submitting AI-generated content as your own work.

• Running documents through an LLM to substitute synonyms or reword sentences without adding clarity.

If you are unsure whether a particular use of AI is appropriate, raise your questions in our Matrix channel or ask the mentor directly in the issue. If you need a quick response, the Matrix channel is a great option to notify us.