LinuxLinks Policy on AI-Heavy Software Projects

Updated 30 September 2026

LinuxLinks reviews and recommends free and open source software. Inclusion on the site is an editorial decision, not an entitlement. An entry in one of our roundups represents a recommendation, so we consider more than a project’s stated features. We also examine its reliability, maturity, maintainability and development practices.

What We Mean by “AI-Heavy”

We use the term “AI-heavy” when generative AI has played a substantial role in producing or modifying a project’s code, tests, documentation, artwork or other important materials.

The term describes how a project was developed, not what the software does. A program that offers AI-related features is not necessarily AI-heavy.

Occasional use of AI for explanations, debugging assistance, code review or routine development tasks does not automatically make a project AI-heavy. There is no dependable way to calculate the precise proportion of a codebase produced by AI, so we do not apply a fixed percentage threshold. Each project is assessed individually.

Our General Position

We do not impose a blanket ban on software developed with the assistance of generative AI. These tools can be useful when employed by developers who understand, verify and take responsibility for the resulting work.

However, substantial reliance on generated material can introduce additional concerns. These include:

• Features that are incomplete or do not work as described.
• Repetitive, unnecessarily complex or poorly structured code.
• Weak error handling and inadequate testing.
• Unsafe, outdated or unnecessary dependencies.
• Unexpected network activity or other security concerns.
• Documentation that does not accurately reflect the software.
• Uncertainty about the origin or licensing of generated material.
• Maintainers being unable to explain, debug or repair their own code.
• Projects being abandoned shortly after their initial release.

For these reasons, AI-heavy projects are generally less likely to receive coverage, particularly when they are new, immature, derivative, poorly tested or presented with exaggerated claims.

Disclosure Requirements

Anyone submitting a project to LinuxLinks must disclose material use of generative AI accurately.

The disclosure should explain, in reasonable terms:

• Which generative AI tools were used.
• Whether they were used for code, tests, documentation, artwork or other project materials.
• The approximate extent of their use.
• How generated material was reviewed, tested and verified by a human maintainer.

For an AI-heavy project to be considered for coverage, we strongly prefer this information to be disclosed prominently in the project’s README, website or development documentation. It should not be hidden in an obscure issue, discussion or commit message.

Failure to disclose substantial AI use, or providing misleading information about it, will normally result in the project being declined.

How Projects Are Assessed

When considering an AI-heavy project, we may examine:

• Whether the software works reliably on Linux.
• Whether its advertised features are actually implemented.
• The quality, clarity and maintainability of the code.
• The project’s testing and release practices.
• The accuracy of its documentation.
• The maintainer’s understanding of the codebase.
• The maintainer’s ability to investigate and correct faults.
• The project’s development history and prospects for continued maintenance.
• Dependency, privacy and security risks.
• The provenance and licensing of project materials.
• Whether the software offers genuine usefulness beyond existing projects.

A repository’s size, commit count or apparent level of activity is not treated as proof of quality or maturity.

Editorial Decisions

We may decline coverage when there is insufficient evidence that a project is reliable, properly maintained or understood by its maintainer. We may also postpone consideration until a project has established a credible development and maintenance record.

Established projects may be treated differently from newer AI-heavy projects. Where a project has a long development history and substantial adoption, we may continue to recommend it even if recent development has become heavily AI-assisted. We distinguish between mature projects that later adopted extensive use of generative AI and newer projects whose codebase has been substantially built through AI-assisted development from an early stage.

An AI-heavy project may still be covered when it is useful, transparent, properly tested and maintained by people who clearly understand and take responsibility for it.

We may revisit existing coverage if a project is abandoned, if serious faults come to light, or if previously undisclosed AI use is discovered. An entry may be amended or removed when the project no longer meets our editorial standards.

We do not maintain a separate category for AI-heavy software. AI use is a development consideration, not a software category, and a separate listing could be interpreted as an endorsement.

Final Decision

All coverage decisions remain at the discretion of LinuxLinks. This policy is intended to protect the quality and usefulness of our recommendations while allowing responsible and transparent use of generative AI in software development.