Solus Linux Sets a Formal Line on AI in Open Source Contributions
Solus Linux has published a formal policy governing how artificial intelligence and large language model tools can be used in project contributions, becoming one of the more explicit Linux distributions to draw a line between acceptable and unacceptable AI-assisted work. The policy reflects a broader tension in open source: how to welcome the productivity gains AI tools offer without eroding the accountability and quality that volunteer-driven projects depend on.
The Rules of the Road
The policy allows developers to use AI to assist with code and package changes, but places several hard requirements on what follows. Every submission must be tested by the contributor before it is committed. If an AI tool was used in producing the work, the commit message must include an "Assisted-by LLM" note. And critically, only registered developers can submit contributions, not automated bots or unreviewed pull requests from unidentified accounts.
The stance is blunt: contributors are expected to understand their own submissions. The Solus team's reasoning is straightforward. The operating system runs on real machines every day, and a bug introduced by AI-generated code that no one fully comprehends is a bug that affects actual users.
The rules grow stricter when it comes to written material. Solus prohibits any AI-generated or partially AI-generated content in its blog posts and official communications. This is a notable divergence from the code policy, suggesting the team views editorial and marketing content as requiring a different standard of authenticity.
Transparency and Consequences
Attempting to conceal the use of AI tools carries real consequences. The policy warns that obscured AI involvement can result in rejected contributions and potential bans from future participation. The team frames the policy as an encouragement of open collaboration rather than surveillance, but the enforcement mechanism makes clear that transparency is not optional.
The "Assisted-by LLM" note is the simplest requirement but also the most consequential one. It creates a visible record that allows other contributors to assess whether a piece of code or text benefited from AI assistance and whether that assistance was applied judiciously.
Where the Policy Came From
Solus is not inventing this framework from scratch. The policy draws directly from the AI contribution guidelines already adopted by Budgie Desktop, the project's default desktop environment. That policy, in turn, takes inspiration from similar efforts by the Linux kernel community and Fedora, both of which have grappled with how to regulate AI-assisted submissions in large collaborative codebases.
The lineage matters because it signals that Solus is aligning with a growing consensus across the Linux ecosystem rather than charting an isolated course. The Linux kernel maintainers have long required contributors to disclose AI assistance, and Fedora has built its own framework around similar principles. Solus is essentially adopting a proven template and adapting it to its smaller project scale.
An Open-Ended Approach
Solus has been candid that the policy is not permanent. The team acknowledges that as AI tools become more integrated into standard development workflows, the rules may need to change. This is a pragmatic admission rather than a loophole: the policy is designed to address the current landscape, where the line between human-written and AI-assisted code is increasingly blurry.
For other open source projects weighing whether to adopt similar rules, Solus offers a middle path that is neither AI-hostile nor AI-permissive without guardrails. It accepts AI as a tool while insisting that human accountability remains non-negotiable. In a space where many projects are still deciding whether to address AI assistance at all, that position is likely to draw attention from maintainers who have been waiting for someone to set a precedent.