# AI Policy

The use of AI assistance, especially in the form of generative AI, continues to be a heated topic in open source communities. We have policies in place regarding the quality of contributions and expectations of contributors to try and maintain a friendly and welcoming developer community:

- No LLM-generated commits (ex: Claude commits). This is automatically enforced by a CI check.
- No LLM-generated communication (ex: issues, PR descriptions, bug reports), except for translation or accessibility purposes.
- No AI-generated art or assets.
- You, as a member of the community, are responsible for every line of code and comment you contribute. This includes all design and implementation decisions.
- You must fully understand your contributions, and be able to hold a conversation about their contents with fellow contributors.
- Contributions should be largely consistent with existing project conventions. Exhibiting multiple clear signs of LLM output is a red flag, but please do not start witch hunts. Some people just have similar writing styles.
- Use of LLMs in design, contributions, or communication should always be disclosed. LLM-powered tab autocompletion for trivial or mechanical edits requires no disclosure.
- Maintainers reserve the right to close any issues or PRs on the premise of being LLM-generated. This is a subjective judgement call made on a case-by-case basis.

In general, we do not police how contributions are made. What matters to us is the quality of contributions and communication surrounding them. This is a human community, and we do not tolerate or wish to interact with raw LLM output. Fellow humans who act in good faith and take accountability for their work are welcome.

Respect these policies when interacting with the community, both in the GitHub repositories and on the Discord server. If you have concerns about these policies or any individuals, please reach out to the maintainers first to discuss them privately.