Linus Torvalds, the creator of Linux and its longtime top maintainer, has delivered a blunt message to programmers who want to ban AI from kernel development: fork it. In a recent exchange on the Linux Kernel Mailing List (LKML), Torvalds made it clear that AI tools are not only welcome but are already an integral part of how the world's largest open-source project operates.

Torvalds was responding to a discussion prompted by a policy statement from the Software Freedom Conservancy (SFC) titled "When Using LLM-backed Generative AI Systems for FOSS Contributions." The SFC suggested that open-source projects should "support, not just tolerate, those who outright reject LLM-gen-AI systems." This touched off a debate among kernel developers, with some wondering if Linux should follow the lead of projects like the Zig programming language, which has adopted strict policies against AI-generated code.

Torvalds left no room for ambiguity. "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," he wrote. "If you don't like it, you can do the open-source thing and fork it."

Why Torvalds Supports AI

Torvalds framed his support in pragmatic terms. "AI is a tool, just like other tools we use. And it's clearly a useful one," he said. "It may not have been that 'clearly' even just a year ago, but it's no longer in question today." He acknowledged that AI tools have flaws and can sometimes cause pain for maintainers, but he argued that the solution is not to reject them outright. Instead, developers should work to make LLM tools more helpful.

This view is rooted in Torvalds' long-standing philosophy for Linux development: decisions should be based on technical merit, not ideology. "In the kernel community, we do open source because it results in better technology, not because of religious reasons," he stated. "And so we make decisions primarily based on technical merit. Not fear of new tools."

The kernel has already been using AI in various ways. AI tools help with code review, bug triage, automated backporting of security patches, and even generating initial patches. Torvalds himself has noted that he is moving away from writing code directly and instead relying on tools that include AI-assisted systems. In a separate interview earlier this year, he said, "I'm not a programmer anymore," referring to how his role has shifted to reviewing and managing pull requests with the help of modern tools.

Other Top Maintainers Agree

Torvalds was not alone in pushing back against anti-AI sentiment. Greg Kroah-Hartman, maintainer of the Linux stable kernel, told ZDNET earlier this year that the quality of AI-generated contributions has dramatically improved. "Months ago, we were getting what we called 'AI slop,' AI-generated security reports that were obviously wrong or low quality," he recalled. "But then the world switched. Now we have real reports. All open-source projects have real reports that are made with AI, but they're good, and they're real."

Kroah-Hartman has also emphasized that AI is a tool that helps maintainers cope with the enormous volume of contributions. The Linux kernel has thousands of contributors and hundreds of patches merged every day. Automated tools, including AI-based ones, are essential to keep the project moving.

Another prominent kernel developer, James Bottomley, weighed in with a clear position on the boundaries of AI rejection: "The contributor doesn't get to approve the tools the maintainer uses to assess and apply patches. If there's AI in there, and the contributor is an AI luddite, then the patch doesn't get applied (i.e., your right to ignore AI stops when it infringes others' right to use it)."

Senior kernel maintainer Theodore "Ted" Ts'o also raised practical concerns about implementing the SFC's suggestion. He asked: "If someone rejects LLM-gen-AI systems, and the LTS kernel contains patches which are automated backported, and they object, are we bound to forswear the use of automated backport technologies? What if someone reports a bug with a kernel stack trace, and someone uses an LLM agent to analyze their bug report and find a fix? What does it mean to 'support someone who outright rejects the use of LLM-gen-AI systems' in that case?" He concluded that it is not obvious that the kernel project must bend over backward to accommodate those who reject AI.

The Broader Open-Source Landscape

The debate over AI in open-source projects is not unique to Linux. Some projects, like the Zig language, have implemented strict policies prohibiting any AI-generated code. The Zig project's maintainer, Andrew Kelley, has argued that AI-generated code lacks the accountability and reliability needed for systems programming and can introduce subtle bugs. He also raised ethical concerns about the energy consumption and data sourcing of large language models.

However, the Linux kernel's stance represents a more dominant view among large open-source projects. Projects like Kubernetes, Node.js, and many others have either embraced AI or taken a neutral position, leaving decisions to individual maintainers. The Apache Software Foundation has issued guidelines for AI contributions but has not banned them.

The SFC's policy was intended to protect developers who do not wish to interact with AI-generated contributions. But in practice, enforcing such a policy becomes incredibly complex. For example, many developers use AI-powered code completion tools like GitHub Copilot or CodeWhisperer. Is a developer who uses such a tool prohibited from contributing? What about using AI to help write a commit message or summarize a bug report? The line is fuzzy.

Torvalds' approach cuts through this confusion: focus on the output, not the tool. If a patch improves the kernel, it doesn't matter whether it was written by a human or assisted by AI. What matters is that it compiles, runs correctly, and abides by the kernel's coding standards.

Historical Context: Torvalds and Controversy

This is far from the first time Torvalds has taken a hard line on a contentious issue. Known for his blunt and sometimes abrasive communication style, he has previously been the subject of controversy over his behavior on the LKML. In 2018, he took a break from kernel maintenance to work on his emotional control and communication. He returned with a promise to be less aggressive, and many have noted a more measured tone in recent years. However, on key technical and policy issues, he remains as firm as ever.

Torvalds has also been a lightning rod in other debates, such as the adoption of the Rust programming language in the kernel. Rust's memory safety features are seen by many as a way to reduce vulnerabilities, but integrating a second language into the kernel has been a massive undertaking. Torvalds has supported this move, praising Rust for making coding fun again, but he has also cautioned that the integration must be handled carefully. The AI debate, however, appears to be one where he sees no need for caution.

The Future of AI in the Linux Kernel

As AI models continue to advance, their role in kernel development is likely to grow. The year 2026 has seen the release of frontier models such as Anthropic's Claude Opus 4.8 and OpenAI's GPT-5, which are increasingly capable of understanding complex codebases and generating high-quality patches. Torvalds noted that the question of whether AI is useful is no longer up for debate. "There are other questions around AI (like what the economy of it will actually look like in the end), but 'is it useful?' is no longer one of those questions," he wrote.

For developers who oppose AI, the message is clear: adapt, fork, or contribute elsewhere. The Linux kernel will continue to use whatever tools help produce better technology. As Torvalds succinctly put it: "I will very loudly ignore people who try to argue against other people from using it."

The kernel's maintainers have spoken with one voice. AI is here to stay, and it is already improving the way the world's most critical operating system is built. Those who cannot accept that have a simple choice to make.

Source: ZDNET News