Linus Torvalds has spoken out with unusual clarity about the role of AI in the development of the Linux kernel. According to the founder of Linux, the project is not opposed to the use of AI, and developers who have fundamental objections to it should not expect the community to change course. In his view, anyone who wants to keep AI out of kernel development would be better off starting their own fork of Linux or leaving the project.
Torvalds made these remarks on the Linux Kernel Mailing List (LKML), where a discussion had arisen about the increasing use of AI tools in developing and maintaining the kernel. As the project’s lead maintainer, he made it clear that, as far as he is concerned, there is no ambiguity on this matter. Linux evaluates new tools based on their technical merit, not on ideological grounds.
According to Torvalds, AI has now become a practical tool for software development. While this might have been a point of debate a year ago, he now sees things differently, reports The Register. Developers who deny the usefulness of AI have, in his view, likely hardly used the technology seriously.
In doing so, Torvalds is explicitly opting for a pragmatic approach. In his view, AI is comparable to other development tools that programmers use on a daily basis. It’s not about whether the technology is new, but whether it contributes to better software.
Change of course
These statements mark a striking change. As recently as late 2024, Torvalds was still very cautious about generative AI. At the time, he described most of the AI hype as marketing and indicated that he would prefer to ignore the subject for the time being.
His current stance is more nuanced. He acknowledges that AI has shortcomings and that the technology can create extra work, for example, because AI systems may uncover errors in the kernel that then need to be investigated. However, he does not see this as a reason to reject AI.
Torvalds emphasizes that kernel developers are not required to use AI tools. At the same time, he does not want others to try to discourage their use. In his view, maintainers should be able to decide for themselves which tools to use, as long as they contribute to the quality of the software.
In doing so, he appears to want to put an end to the principled debate about AI within the Linux community. He believes the focus should be on the technical value of these tools, not on resistance to new technology.
AI output improved
These remarks are in line with earlier signals from the Linux community. Greg Kroah-Hartman, also a prominent kernel maintainer, said earlier this year that AI-generated bug reports and assistance with code reviews have noticeably improved in recent months. While such reports were often unusable at first, he says they now regularly provide valuable information.
At the same time, Torvalds acknowledges that AI is far from flawless. Still, he puts that criticism into perspective. After all, human developers make mistakes too, which is why, in his view, AI shouldn’t be the only one held to that standard.
Torvalds’ message is thus clear: AI is not a mandatory part of the development process, but it is a technology that has earned a legitimate place within the Linux project.