Jellyfin, a popular open-source media server, has released a new LLM development policy. The policy addresses the rising use of AI tools like ChatGPT in its codebase. The project emphasizes code quality, readability, and simplicity. This move reflects a growing concern within the developer community regarding the integration of AI-generated code.

The policy prohibits direct use of LLM output for communication, including comments and pull request descriptions. For code contributions, Jellyfin requires developers to understand and explain changes made by LLMs. Any LLM-assisted code must adhere to strict formatting and quality standards. This is to ensure the integrity of the project's codebase and to maintain its core values.

Developers must also be able to review the output and explain the reasoning behind changes. The policy also covers LLM-generated tools shared within the community. Projects must be clearly marked as LLM-generated. Jellyfin's moderators won't actively police third-party projects, but will enforce rules against code theft and bad-faith attribution.

This policy reflects a broader debate about the impact of AI on software development. As AI tools become more prevalent, projects like Jellyfin are establishing guidelines to balance the benefits of AI with the need for quality, maintainability, and community standards. The final discretion always lies with the reviewers.