Debating the role of large language models in the kernel community [LWN.net]
LWN<br>.net<br>News from the source
Content Weekly Edition<br>Archives<br>Search<br>Kernel<br>Security<br>Events calendar<br>Unread comments
LWN FAQ<br>Write for us
Edition Return to the Front page
User:<br>Password: |
Log in /<br>Subscribe /<br>Register
Debating the role of large language models in the kernel community
Did you know...?
LWN.net is a subscriber-supported publication; we rely on subscribers<br>to keep the entire operation going. Please help out by buying a subscription and keeping LWN on the<br>net.
By Jonathan Corbet<br>July 21, 2026
Like many development communities, the kernel community has been struggling<br>to determine how large language models (LLMs) will be used in its development<br>process. The news has been dominated recently by a strongly worded missive<br>from Linus Torvalds on the subject, but the discussion has been rather more<br>wide-ranging and nuanced than that. Topics that have been considered<br>recently include the LLM attribution requirement, code-review tools,<br>dependence on proprietary tools, and whether there is a place for concerns<br>about the ethics of LLMs.
Reconsidering Assisted-by
The kernel community's guidance regarding LLM-generated code was added to the kernel, at the<br>end of 2025 for the 7.0 release, after long discussions in the Linux<br>Foundation's Technical Advisory Board (TAB) and on the mailing lists. The<br>coding-assistants.rst<br>file states that, when a code submission has been partially or entirely<br>generated by an LLM, the patch should carry an Assisted-by tag in this<br>format:
Assisted-by: AGENT_NAME:MODEL_VERSION [TOOL1] [TOOL2]
The purpose of this tag was to document the use of an LLM, and to alert<br>reviewers and maintainers to that fact. It also contains the name of the<br>specific tool used. That information was deemed potentially useful in case<br>a specific LLM turns out to have a tendency toward specific bugs or ends up<br>in copyright-related trouble; it would make it possible for developers to<br>review patches generated by the problematic model and, possibly, modify or<br>remove them.
Since then, this tag has not won over all developers. As of 7.2-rc4, over<br>1,200 commits carry Assisted-by tags. There is, however, clearly a<br>significant stream of machine-generated patches that do not carry that tag;<br>sometimes that is a result of ignorance of the rules, but other times the<br>origin of the code is, seemingly, being deliberately obscured. As a<br>result, the tag's usefulness as an indicator of LLM involvement is unclear<br>at best. Meanwhile, some developers see placing the names of specific LLMs<br>into the kernel's development history as a form of advertising. All told, many<br>wonder whether the tag adds any value at all.
This discussion came to the fore at the beginning of July when Christian<br>Brauner suggested<br>removing the tag or, at least, removing the name of the specific model<br>used:
I acknowledge that my stance is even more radical: imho we would<br>just stop it with any disclosure requirements completely. It's<br>useless imho. We already see that other than core contributors<br>most people don't care and will just not disclose their usage of<br>AI. I think this is entirely pointless and worse it brings in<br>undefined legal status as well. It's not like recent events of<br>pulling certain models from the face of the earth have made this<br>any less concerning.
But fine, if we want to do this can we please just dumb it down to
Assisted-by: LLM
Jeff Layton later followed this up with a separate<br>patch removing the attribution requirement entirely. Some developers<br>were clearly in favor of that; networking maintainer Jakub Kicinski let it be known that<br>he simply removes those tags from patches he applies, essentially<br>undermining the previous decision made by the TAB and the community.<br>Others have not taken that step, but expressed little love for the tag.
There are developers, though, who would like to see its use continue. Greg<br>Kroah-Hartman described it as a<br>signal that an LLM was involved in the creation of a patch, and that the<br>result should be reviewed more closely. Lorenzo Stoakes characterized the current<br>policy as being "about empowering maintainers to push back". This<br>contingent would like to see the Assisted-by requirement stay in force,<br>perhaps with better enforcement if possible.
In the end, there appeared to be enough support for keeping the tag to<br>prevent its outright removal, but nobody was willing to defend requiring<br>disclosure of the specific tool used. So the consensus seemed to be<br>to go with Brauner's initial suggestion, perhaps with a more focused<br>changelog. A new revision has not yet appeared, though, and it may well be<br>that a final decision on this policy will not be made until the Maintainers<br>Summit in October.
Dependence and sustainability
While the use of LLMs to write code is clearly on the increase in the<br>community, LLMs are still much more widely used for the review of patches<br>rather than their creation. The Sashiko<br>review tool, in...