The Fedora Linux project has formally approved a new policy allowing AI-assisted contributions to its codebase. Announced by Aoife Moloney on the Fedora discussion board, the policy emphasizes accountability and transparency while setting boundaries on AI use. This development follows similar discussions in related open-source projects like Mesa.
The Fedora Council recently approved the latest version of the AI-Assisted Contributions policy, as announced by Aoife Moloney on the Fedora discussion board. This policy permits contributors to use AI tools for Fedora projects, provided they adhere to specific principles outlined in the document.
Under the policy, titled 'Fedora AI-Assisted Contributions Policy,' key principles include:
Accountability: Contributors must take full responsibility for their submissions, vouching for quality, license compliance, and utility. The policy states, 'All contributions, whether from a human author or assisted by large language models (LLMs) or other generative AI tools, must meet the project’s standards for inclusion. The contributor is always the author and is fully accountable for the entirety of these contributions.'
Transparency: Significant use of AI without changes must be disclosed, while other uses should be noted where helpful. Routine tasks like grammar and spelling corrections do not require disclosure. For git-tracked contributions, disclosures use an 'Assisted-by:' commit message trailer, such as 'Assisted-by: generic LLM chatbot' or 'Assisted-by: ChatGPTv5.'
Contribution & Community Evaluation: AI may assist human reviewers with analysis and suggestions but cannot serve as the sole arbiter for substantive judgments or community standing decisions. Automated tools for objective validation, like CI/CD pipelines, remain allowed, with final accountability on humans.
The policy excludes large-scale initiatives that could alter project operations, requiring separate discussions with the Fedora Council. Concerns about violations should be reported via private tickets to the Council.
This approval aligns with a recent update to the Mesa graphics drivers contributor guidelines, which mentions AI, though debates on allowing such contributions in Mesa continue.