Linus Torvalds weighs in on AI contribution rules
| Source: HN | Original article
Linus Torvalds outlines his stance on rules governing AI contributions to open‑source projects.
Linus Torvalds has reiterated his stance on the use of artificial‑intelligence tools in Linux kernel development, emphasizing that any contributions generated by AI must be clearly identified and vetted by a human maintainer before they are accepted. The kernel maintainer’s comments, made during a recent community forum, echo his earlier warnings that “blindly trusting AI‑generated code is a recipe for bugs and regressions.”
Torvalds’ position matters because the Linux kernel remains the backbone of countless devices and services worldwide, and its development model relies on rigorous peer review. As large language models become increasingly capable of producing code snippets, the risk of introducing subtle defects or security flaws grows. By insisting on explicit disclosure and human oversight, Torvalds aims to preserve the kernel’s reliability while still allowing developers to experiment with AI assistance where it adds genuine value.
The remarks come after a series of high‑profile experiments in which Torvalds permitted Claude Code to draft patches for the page‑cache subsystem and other components. Those trials, reported in our coverage on 15 October 2026, revealed both the promise of rapid prototyping and the pitfalls of race conditions that required manual correction. Earlier, on 13 October 2026, we noted his criticism of “no‑AI” policies as insufficient safeguards, calling them “Level 0 controls.” The current clarification builds on that dialogue, suggesting a middle ground: AI may be used, but only under strict attribution and review protocols.
What to watch next is how the kernel community translates Torvalds’ guidance into concrete contribution guidelines. The upcoming 6.9 release cycle will likely include a formal amendment to the kernel’s “Signed‑off‑by” process, and several major subsystem maintainers have signalled they will draft their own AI‑contribution policies. Observers will also be keen to see whether other open‑source projects adopt similar rules, potentially shaping a broader industry standard for responsible AI‑assisted development.
Sources
Back to AIPULSEN