The Linux kernel community has agreed to eliminate the requirement that contributors explicitly label patches created with artificial intelligence tools.

The change emerged from a discussion among maintainers about the practicality of tracking AI-assisted contributions. Sasha Levin, a Linux kernel developer affiliated with Nvidia, announced that the project will no longer mandate the "Assisted-by" tag in commit messages, stating that the documentation requiring such disclosures is being deprecated.

The policy had previously instructed developers using any advanced coding tool to indicate its use through a specific tag. This rule was formally introduced in January 2026, following earlier discussions that began in July 2025. Its development coincided with broader industry conversations about AI's role in software engineering, including statements from Linus Torvalds, who had expressed cautious openness to AI tools while emphasizing the importance of code maintainability.

Greg Kroah-Hartman, a long-time Linux maintainer, noted at a recent summit that modern AI systems are now capable of producing code of sufficient quality for kernel development, particularly in identifying defects through pattern recognition. He observed that AI-generated code has reached a level of reliability that warrants its integration into established workflows, though concerns about error detection and accountability remain.

Despite this shift, some contributors continue to advocate for transparency measures. Laurent Pinchart, a kernel developer, argued that while the tagging system may be cumbersome, it serves a valuable purpose in setting clear expectations and maintaining accountability. He warned that removing the requirement without an alternative disclosure mechanism could weaken oversight, as some developers might otherwise avoid responsibility for AI-generated content.

Torvalds responded strongly to these concerns, dismissing the notion that AI tool usage should be restricted or singled out. He criticized the idea of imposing special rules for AI while allowing other development tools to operate without similar scrutiny. In a public message, he questioned the logic of such distinctions, suggesting that an overemphasis on tool provenance risks undermining practical development standards. He emphasized that understanding code is more important than the tools used to produce it, and that experienced contributors should be trusted to evaluate contributions on their merits rather than their origins.

The debate reflects broader tensions in the open-source community as AI tools become more embedded in software creation. While some view AI as a natural evolution of development practices, others remain cautious about its impact on code quality, security, and the long-term sustainability of collaborative projects. The removal of the disclosure rule marks a significant moment in this evolving landscape, signaling a pragmatic approach to integrating new technologies without compromising established norms.