Torvalds backs AI-assisted kernel work that meets Linux standards
Linus Torvalds supports competent developers using AI to help write Linux kernel code, arguing that the quality of the work matters more than whether AI was involved. The article says kernel maintainers have dropped the requirement to disclose AI assistance, while stressing that submissions still need to meet the project’s standards.
At a Linux Kernel Maintainers Summit in Prague, developer Sasha Levin announced the decision to deprecate the “Assisted-by” tag. Torvalds distinguishes experienced developers using AI as a tool from “drive-by” contributors who submit patches without learning or engaging with the community. The article argues AI is becoming part of software development, while acknowledging concerns about model reliability, copyright and low-quality AI output.
In short
- Torvalds backs competent developers using AI for kernel code.
- Maintainers no longer require an AI assistance tag.
- Low-quality “drive-by” patches still face rejection.
Steelman: both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
If experienced developers use AI as a tool while taking responsibility for the result, judging patches by their quality and compliance with Linux standards keeps review focused on what matters. Dropping a disclosure tag may also avoid singling out one tool or process as AI becomes a routine part of software development.
The case against
Disclosure can help maintainers understand how a patch was produced and scrutinise it appropriately, especially given concerns about model reliability, copyright and low-quality output. Keeping an “Assisted-by” tag could support transparency and help the community assess AI’s effects without preventing capable developers from using it.
Explained simply
The Linux kernel is the core software that lets the Linux operating system communicate with computer hardware. It is developed in the open by contributors around the world, with maintainers reviewing proposed changes before they are included.
Linus Torvalds created Linux and remains a leading figure in its development. Maintainers such as Sasha Levin help oversee contributions, while developers submit patches—changes to the code—for review. The project’s standards and review process are intended to keep the kernel reliable.
AI tools can help developers write or check code, but their output can contain errors. The debate matters because the kernel is widely used, and its contributors must decide how to assess code as AI becomes more common in software work.
Read the full article at the source →
Originally published by The Register as “Torvalds: AI is fine for competent devs, but spare us the ‘drive-by’ patches”.