Sashiko: AI code review system for the Linux kernel spots bugs humans miss
(2026/03/20)
- Reference: 1773999007
- News link: https://www.theregister.co.uk/2026/03/20/sashiko_code_review_linux/
- Source link:
AI is coming to the Linux kernel in the form of a code review system - not code submissions.
[1]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.
Gushchin said: "In my measurement, [2]Sashiko was able to find 53 percent 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 percent is not that impressive, but 100 percent of these issues were missed by human reviewers."
[3]
The use of AI in code submissions is [4]controversial in the open source community, however, a tool like Sashiko could go some way toward easing the burden on maintainers dealing with a wave of code reviews.
[5]Systemd 260 kills SysV, tells AI not to misbehave
[6]Linux Foundation kicks off effort to shield FOSS maintainers from AI slop bug reports
[7]Bcachefs creator insists his custom LLM is female and 'fully conscious'
[8]Godot maintainers struggle with 'draining and demoralizing' AI slop submissions
Sashiko works by ingesting patches from a mailing list. It analyzes them then gives feedback to the maintainers and developers. According to its authors, "the quality of reviews is high... the rate of false positives is harder to measure, but based on limited manual reviews it's well within 20 percent range, and the majority of it is a gray zone."
The authors are upfront about the privacy and code-sharing aspects. 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. However, there are costs involved in running it. In the case of the Linux Kernel Mailing List, Google is footing the bill.
[9]
Gushchin said: "We've been using it internally at Google for some time, and it helped to discover a large number of real issues."
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. ®
Get our [10]Tech Resources
[1] https://www.linkedin.com/posts/roman-gushchin-1a95658_im-happy-to-share-the-project-i-was-working-activity-7439736656360005632-M-cb
[2] https://github.com/sashiko-dev/sashiko
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ab0o0DYNZFHy2KAPcReXMgAAAA8&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[4] https://www.theregister.com/2026/02/03/github_kill_switch_pull_requests_ai/
[5] https://www.theregister.com/2026/03/18/systemd_260/
[6] https://www.theregister.com/2026/03/18/linux_foundation_ai_slop_defense/
[7] https://www.theregister.com/2026/02/25/bcachefs_creator_ai/
[8] https://www.theregister.com/2026/02/18/godot_maintainers_struggle_with_draining/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ab0o0DYNZFHy2KAPcReXMgAAAA8&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/
[1]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.
Gushchin said: "In my measurement, [2]Sashiko was able to find 53 percent 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 percent is not that impressive, but 100 percent of these issues were missed by human reviewers."
[3]
The use of AI in code submissions is [4]controversial in the open source community, however, a tool like Sashiko could go some way toward easing the burden on maintainers dealing with a wave of code reviews.
[5]Systemd 260 kills SysV, tells AI not to misbehave
[6]Linux Foundation kicks off effort to shield FOSS maintainers from AI slop bug reports
[7]Bcachefs creator insists his custom LLM is female and 'fully conscious'
[8]Godot maintainers struggle with 'draining and demoralizing' AI slop submissions
Sashiko works by ingesting patches from a mailing list. It analyzes them then gives feedback to the maintainers and developers. According to its authors, "the quality of reviews is high... the rate of false positives is harder to measure, but based on limited manual reviews it's well within 20 percent range, and the majority of it is a gray zone."
The authors are upfront about the privacy and code-sharing aspects. 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. However, there are costs involved in running it. In the case of the Linux Kernel Mailing List, Google is footing the bill.
[9]
Gushchin said: "We've been using it internally at Google for some time, and it helped to discover a large number of real issues."
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. ®
Get our [10]Tech Resources
[1] https://www.linkedin.com/posts/roman-gushchin-1a95658_im-happy-to-share-the-project-i-was-working-activity-7439736656360005632-M-cb
[2] https://github.com/sashiko-dev/sashiko
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2ab0o0DYNZFHy2KAPcReXMgAAAA8&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[4] https://www.theregister.com/2026/02/03/github_kill_switch_pull_requests_ai/
[5] https://www.theregister.com/2026/03/18/systemd_260/
[6] https://www.theregister.com/2026/03/18/linux_foundation_ai_slop_defense/
[7] https://www.theregister.com/2026/02/25/bcachefs_creator_ai/
[8] https://www.theregister.com/2026/02/18/godot_maintainers_struggle_with_draining/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44ab0o0DYNZFHy2KAPcReXMgAAAA8&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/