Google engineers have developed Sashiko, an agentic AI code review system for the Linux kernel that has been used internally at Google for some time to uncover issues and is now publicly available. Announced on LinkedIn by Roman Gushchin, a Linux kernel engineer at Google, Sashiko is a tool written in Rust for spotting bugs and screening code.
The numbers are striking. Sashiko was able to find 53 per cent of bugs based on a completely unfiltered set of 1,000 recent upstream issues based on 'Fixes:' tags using Gemini 3.1 Pro. Some might say that 53 per cent is not that impressive, but 100 per cent of these issues were missed by human reviewers.
The maintainer crisis
This matters because the Linux kernel is one of the largest and most active collaborative software projects on Earth. Thousands of patches flow through the kernel mailing lists every month. Maintainers, many of whom are already stretched thin, must review each one for correctness, style, security implications, and compatibility with an enormous and growing codebase. Burnout among kernel maintainers isn't a hypothetical risk. It's a documented, ongoing crisis.
Sashiko operates as a first-pass screening tool. Sashiko works by ingesting patches from a mailing list. It analyzes them then gives feedback to the maintainers and developers. The system does not merge or reject code; humans retain the final decision.
How it works
Sashiko uses a multi-stage review protocol to evaluate patches thoroughly from multiple perspectives, mimicking a team of specialised reviewers. It uses a set Linux kernel-specific prompts and a special protocol to review proposed Linux kernel changes.
Sashiko sends data and code to whatever LLM provider it has been configured for. It has been most tested with Gemini Pro 3.1, but should work with Claude and other LLMs. In the case of the Linux Kernel Mailing List, Google is footing the bill. This arrangement sidesteps the cost barrier that would otherwise make such a system impractical for volunteer-run projects.
A different approach to AI in development
What makes Sashiko distinct is that it serves as an advisor, not a code generator. Unlike the controversy surrounding AI that writes code directly, this tool works within a review framework humans have used for decades. Think of it as an automated code reviewer that can identify issues ranging from formatting problems and coding style violations to more substantive concerns like potential memory safety bugs or logic errors.
Yet concerns linger. The Linux kernel community has a long and sometimes contentious history with automation in the development process. Some developers worry that the kernel development culture is famously rigorous, sometimes abrasive, and deeply protective of code quality. Introducing AI into this environment requires navigating not just technical challenges but deeply held cultural norms about how software should be written, reviewed, and maintained.
Sashiko belongs to the Linux Foundation and looks like a useful tool – an application of agentic AI that may provoke less handwringing than code submissions. For now, the experiment proceeds with cautious interest. The kernel needs the relief. What remains to be seen is whether the community will accept help that arrives without a human reputation behind it.