Linus Torvalds has drawn a firm line in the open source debate over artificial intelligence, declaring that the Linux kernel is not an anti-AI project and telling developers who object to AI-assisted coding that they are free to fork the project or leave.
“Linux is not one of those anti-AI projects, and if somebody has issues with that, they can do the open-source thing and fork it. Or just walk away,” Torvalds wrote in a post on the Linux Kernel Mailing List on July 15. “I realize that some people really dislike AI, but this is an area where I’m willing to absolutely put my foot down as the top-level maintainer.”
The statement marks a significant evolution for Torvalds, who in October 2024 described nearly 90 percent of AI messaging as “marketing hype.” By mid-2026, he has become an active defender of AI tools in kernel development.
“AI is a tool, just like other tools we use,” Torvalds wrote. “And it’s clearly a useful one. It may not have been that clearly even just a year ago, but it’s no longer in question today. Anybody who doubts that clearly hasn’t actually used it.”
He added: “We’re not forcing anybody to use it, but I will very loudly ignore people who try to argue against other people from using it.”
The immediate trigger was a discussion around Sashiko, an AI-powered patch review system for the Linux kernel developed by Google engineer Roman Gushchin. Operating under the Linux Foundation, Sashiko uses large language model agents to automatically scan kernel submissions and flag potential issues before human maintainers review them. The tool has reportedly identified around 53 percent of bugs in tested code sets since becoming open source in March 2026.
The kernel’s current policy permits AI-assisted development but holds human contributors responsible for reviewing, disclosing, and signing off on the resulting work through the Developer Certificate of Origin (DCO). Some open source projects have prohibited AI-assisted contributions altogether, citing concerns about low-quality automated submissions, increased review burdens, and uncertain code provenance.
Torvalds is not blind to the downsides. In May 2026 he noted that AI-generated bug reports had made the kernel security mailing list “almost entirely unmanageable,” and acknowledged that LLM outputs can be inaccurate. But his position is that ideological bans on useful technology are counterproductive.
Sources: Ars Technica; Virtualization Review