Fedora Council approves AI-assisted contributions policy (discussion.fedoraproject.org)

The Fedora Council approved a permissive AI-assisted contributions policy that deliberately avoids hard definitions of “contribution” or “contributor.” Rather than codify narrow categories like git patches, the policy casts a wide net to include code, docs, blogs, outreach, design assets and other community work, leaving judgement to existing human reviewers and community intuition. Leaders noted this ties into Strategy2028 (which aims to double weekly active contributors) and that a Fedora Data Working Group is being created to better measure and categorize contributions over time. Technically and operationally, the policy emphasizes human oversight: most release notes, change proposals and marketing content already go through substantial peer review, so moderate AI use is acceptable provided community review persists. Maintainers’ RPM spec changelogs are expected to be little affected—AI could even help generate more thorough changelogs where they’re currently sparse. The Council also set a practical feedback loop: if “AI slop” appears in a particular area within about three months, the policy will be revisited and can be tightened for specific contribution types. Overall, this is a flexible, community-centric approach that enables AI-assisted productivity while preserving review processes and allowing data-driven refinement.