Linus Torvalds has strongly criticized mandatory AI usage disclosure rules in the Linux kernel development process as 'utterly stupid,' joining maintainers who decided to drop the 'Assisted-by' tag that required contributors to declare AI tools in their code commits.
- Linux removes 'Assisted-by' tag that mandated AI usage disclosure in code commits.
- Torvalds calls mandatory AI reporting rules 'utterly stupid' and emphasizes skill over tools.
- Debate reflects broader open source tensions around AI, transparency, and code quality.
What happened
The Linux kernel maintainers have officially deprecated the 'Assisted-by' tag, a requirement introduced in January for contributors to disclose use of AI or advanced coding tools when submitting code. The decision followed a proposal by Nvidia engineer and kernel developer Sasha Levin to remove the documentation and mandate for such disclosure, a move that garnered over 50 replies on the kernel mailing list.
Torvalds publicly pushed back on AI-related disclosure policies in a direct exchange on the mailing list, dismissing the idea of mandatory markers as 'utterly stupid.' Other kernel developers voiced concerns that the tag was inconsistently applied and inadvertently promoted proprietary AI companies. Despite objections calling for continued disclosure to maintain contribution quality expectations, maintainers reached consensus to remove the tag.
Why it matters
This change reflects growing tensions within the open source community about balancing transparency, trust, and productivity in an era where AI-assisted coding becomes more prevalent. While some developers see disclosure as a way to maintain accountability and code quality, others, including Torvalds, argue that a contributor's skill and understanding are paramount, regardless of the tools used.
The debate highlights broader challenges in regulating AI's role in software development, particularly in critical projects like the Linux kernel. With AI tools increasingly finding security bugs and optimizing legacy code, maintainers face pressure to leverage these advancements while managing the related stresses and ethical considerations around AI-generated contributions.
What to watch next
Open source projects may continue to wrestle with how to integrate AI assistance transparently and fairly without hampering innovation or placing undue burdens on maintainers. The Linux kernel's experience could influence policies in other significant projects, especially as AI tools improve and become ubiquitous in programming workflows.
Additionally, monitoring how maintainers address the downstream impact of AI — including review automation with systems like Sashiko — and how communities handle AI-generated content bans, as seen in System76’s COSMIC and NetBSD projects, will offer insights into evolving norms and governance around AI-assisted open source development.