Linux kernel drops AI-assistance tag requirement, Hackaday reports

Linux kernel maintainers have decided to drop the requirement to mark commits made with help from large language models, Hackaday reports. The change puts a major open-source project on a more permissive path for AI-assisted work, while leaving the question of how that work is reviewed firmly on the table.

Watch Desk analysis

What happened

Hackaday says the kernel project is removing its requirement for an “Assisted-by” tag on LLM-assisted commits. The report also describes Sashiko, a bot used for code review, and says budget has been secured for Meta’s models to review patches and help create or edit commit messages. It says multiple frontier models are run on each patch to try to catch confabulations.

The report characterises the tagging decision as not fully uncontroversial, while saying it appears certain to proceed. That is a reported account of the project’s direction, not a claim that every maintainer agrees.

Why it matters

A tag gives reviewers and future maintainers a way to see where AI assistance entered the process. Dropping that requirement changes what can be inferred from a commit’s record, even as AI tools take on more review and drafting work. In a project as widely used as Linux, the balance between encouraging useful assistance and preserving a legible history matters well beyond one repository.

Our read

This is a consequential experiment in trust, not proof that AI review is either safe or useless. Running multiple models may help catch mistakes, but the key test is whether human maintainers can still understand, challenge and take responsibility for the code that lands. The tag debate is a reminder that transparency is part of the tooling, not decorative paperwork.

What to watch

• Whether the tagging requirement is formally removed and how the change is documented.
• How maintainers describe the role of Sashiko and Meta’s models in patch review.
• Whether the project shares evidence about errors caught or introduced by AI-assisted review.

Discussion spark: Should open-source projects require contributors to disclose AI assistance, or is effective human review enough?

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.