Linus Torvalds, creator of the Linux kernel, has defended the use of AI-powered coding tools within the open source project, stating that those who oppose them should fork the project or leave. In a recent post on the Linux kernel mailing list, Torvalds emphasized that Linux is not an anti-AI project and that users who disagree can choose to fork the code or opt out entirely. His comments came amid a debate over the use of Sashiko, an AI-driven code review system that claims to find 53.6 percent of bugs that would later be fixed by human coders. However, the tool also generates false positives at a rate estimated to be within the 20% range. Torvalds said that while AI isn't perfect, it's a useful tool like any other, and he urged critics to compare its output to that of human maintainers. He argued that natural intelligence isn't always reliable either, and that the benefits of AI should be judged on technical merit rather than fear of new tools. Last year, a study found that open source coders using AI tools were 19 percent less productive than those who didn't, but a follow-up update suggested that developers may now be more sped up by AI in early 2026 compared to early 2025. Torvalds, who has been involved with Linux since its creation in 1991, has experimented with AI tools for personal projects, such as creating a Python audio visualizer for a guitar pedal. However, not everyone in the open source community supports AI coding tools, as seen in a recent incident where a developer introduced a malicious prompt-injection instruction to disrupt AI-assisted coding in the jqwik Java testing library.
Torvalds said that he is not forcing anyone to use AI tools, but he will not tolerate those who try to argue against others using them. He emphasized that AI is a tool, much like other tools developers use, and that its usefulness is now clear. He acknowledged that AI isn't perfect but urged critics to look at the performance of human coders as well. Torvalds also noted that his own experimentation with AI tools has shown their potential, even if the results are not always flawless. In a January post, he described using an AI tool to create a Python audio visualizer for a hobbyist project, which he said started as a typical 'google and do the monkey-see-monkey-do' approach but evolved into a more direct use of AI. Despite the growing adoption of AI in open source development, Torvalds remains pragmatic, focusing on the technical merits of these tools rather than the fear they may inspire.
The debate over AI in open source development highlights the broader tension between those who embrace new technologies and those who resist them. Torvalds’ stance reflects a growing acceptance of AI as a valuable asset in software development, even as concerns about its reliability and impact on productivity persist. His comments underscore the importance of balancing innovation with caution, ensuring that tools like AI are used effectively without compromising the integrity of open source projects.
Source: arstechnica
