Back
Linux kernel ends required AI disclosure tag for contributions
SiTech AI Team2 min read

Linux kernel ends required AI disclosure tag for contributions

The Linux kernel community has agreed to deprecate the Assisted-by tag that once required patches to disclose AI or advanced coding tool use. Maintainers say contributors must still understand submitted code.

Kernel ends AI tag requirement

The Maintainers Summit reached consensus to deprecate the Assisted-by tag, Sasha Levin said in a Linux kernel mailing list message, and the documentation requiring it should be removed. The policy had told contributors using any sort of advanced coding tool in creating a patch to acknowledge it in the commit message.

The AI labeling proposal took shape in July 2025 and became part of the documentation in January 2026, shortly after Linus Torvalds said he was fairly positive about vibe coding as a way into programming while remaining skeptical about maintainability. Levin is a Linux kernel developer and engineer at Nvidia.

Quality changes the argument

AI attribution was less of an issue when generated code was obvious, buggy, and bad, but Linux kernel maintainers say coding agents have in the past year or so produced code of acceptable quality for the community. Greg Kroah-Hartman said at the Linux Foundation Open Source Summit in South Korea in August that LLMs are good enough now and can find bugs because they act as pattern matchers on code.

Concerns remain about errors slipping through unnoticed, accountability, and coding agents reproducing copyrighted code. Jeff Layton, a Linux kernel developer with Red Hat, said the tag was not universally followed, acted as free advertising for proprietary LLM companies, and its purpose was unclear.

Disclosure still divides contributors

Laurent Pinchart, a Linux kernel developer and founder of Ideas on Board, said he was not happy removing the tag without another required disclosure mechanism and could foresee submitters pushing back on questions. He argued the issue was setting expectations, and that some people will not care or cheat while those contributions get rejected.

Torvalds rejected tool usage declarations, writing that expectations should focus on people doing their best and understanding what they are doing, not on which tools they use. He compared demands against AI tool use to banning compilers, assembler only coding, or editors for commit messages, calling such rules stupid.

Sources: The Register

SSiTech

SiTech — AI-powered web development

We build fast, modern websites and bring AI into real business workflows. Have a project or a question? We'd love to help.