- Updated: March 18, 2026
- 2 min read
Google’s Sashiko Agentic AI Code Review Accelerates Linux Kernel Development
Google unveils Sashiko – an agentic AI for Linux kernel code review
Google has announced Sashiko, an advanced agentic AI system designed to automate code review for the Linux kernel. The project, detailed in a recent Phoronix article, showcases how machine learning can assist developers by detecting bugs, style issues, and potential regressions before code lands upstream.
Why Sashiko matters for the Linux community
The Linux kernel is one of the world’s most complex open‑source projects, with thousands of contributors and millions of lines of code. Traditional manual reviews are time‑consuming and prone to human error. Sashiko aims to reduce review latency, improve code quality, and free maintainers to focus on higher‑level design decisions.
Key capabilities and performance
- Automated detection of common C‑language pitfalls and kernel‑specific patterns.
- Integration with existing Gerrit workflows, providing inline comments and suggestions.
- Benchmarks show a 30‑40% reduction in review turnaround time for test patches.
Google engineers’ perspective
According to statements from the Sashiko team, the AI model was trained on years of kernel commit history and leverages Google’s internal AI infrastructure. Engineers emphasize that Sashiko is not a replacement for human reviewers but a complementary tool that surfaces obvious issues early.
Open‑source release and community involvement
Google plans to release Sashiko under an open‑source license later this year, inviting kernel maintainers to contribute to its evolution. The project will be hosted on GitHub, with documentation and contribution guides.
Read more on UBOS Tech
For deeper insights into AI‑driven development tools, explore our related guides:
Stay tuned as Sashiko matures and reshapes the future of collaborative kernel development.
Andrii Bidochko
CTO UBOS
Andrii Bidochko is an AI entrepreneur and researcher focused on AI agents, reinforcement learning, and autonomous systems. He writes about the technologies shaping the future of machine intelligence, from frontier models and agent architectures to real-world AI applications.