Linus Torvalds has rejected a proposal to ban AI tools from Linux kernel development, telling dissenters they are free to fork the project or leave. Instead, he argues the community should adapt LLMs to help maintainers while minimizing incorrect patches and bug reports. Current rules allow AI-generated code with an Assisted-by tag, and since a model cannot sign a certificate of origin, the human submitter bears full responsibility for quality, code review, and legal risk. Torvalds emphasized that open-source philosophy is about building technology, not enforcing ideologies — patches will continue to be accepted solely on technical merit, regardless of the tools used to produce them.

Linus Torvalds on AI in kernel development