AIThis post was created with the assistance of artificial intelligence (AI).

TL;DR

Linus Torvalds has said he uses and supports AI coding tools, but has also stressed that Linux kernel submissions must be work contributors understand and stand behind. His position leaves tool choice open while keeping code quality, rather than whether AI was used, as the standard for acceptance.

Linus Torvalds has defended the use of AI coding tools, including for Linux kernel work, while saying contributors remain responsible for understanding and producing quality submissions. The position matters to the Linux development community as maintainers debate how to handle AI-assisted code and whether contributors should have to disclose its use.

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

In a report published October 26, XDA Developers said Torvalds had described using a large language model to help locate and fix a bug, and had allowed it to draft a commit message. The report also says he used AI to write a kernel patch. These examples show that he is open to the tools; they do not mean every AI-generated patch is suitable for inclusion.

The debate has also concerned disclosure. As Linux maintainers moved to remove a requirement that contributors declare LLM use in submissions, Torvalds objected to creating rules that prescribe which tools people may use. XDA reported that he called the proposed approach “utterly stupid” and argued that contributors should be expected to do their best and understand their work.

Torvalds has also spoken positively about AI-assisted programming for newcomers and personal projects. But his support for experimenting with AI is not the same as approval for unreviewed or poorly understood code in the kernel. The practical distinction in his position is between using a tool and meeting the project’s existing expectations for a submission.

At a glance
analysisWhen: Published October 26, 2026; reflects st…
The developmentXDA Developers reports that Linus Torvalds supports using AI tools in software development, including kernel work, without endorsing lower standards for Linux submissions.
Could Linus Torvalds’ AI Approach Be A Healthier Way To Work With The Tech?

Open source · AI & code · At a glance

Could Linus Torvalds’ AI Approach Be A Healthier Way To Work With The Tech?

Linus Torvalds supports experimenting with AI coding tools while keeping responsibility for every submission with its contributor. For Linux kernel work, the measure is whether a patch is understood, reviewable, and up to standard.

Report date

Oct 262026 · XDA Developers

AI stance

OpenUse tools appropriately

Kernel bar

QualityReviewable, understood work

Policy status

UnclearFinal disclosure details not given

01 / What the report says

Tool use is not a shortcut past review

XDA describes Torvalds using AI in bug fixing and kernel-related work. Those examples signal openness to the tools, not automatic approval for generated patches.

01 · Debugging

Finding and fixing a bug

A large language model reportedly helped Torvalds locate and fix a bug, showing one practical use for AI in development.

02 · Communication

Drafting a commit message

The report says he also let AI draft a commit message: assistance with a task, with the contributor still accountable.

03 · Kernel work

Writing a patch

XDA reports AI helped write a kernel patch. Any contribution still needs to meet Linux project standards and review.

02 / The review standard

Two questions, one acceptance bar

The reported position centers the submitted work and its ownership, rather than making the tool itself the measure of quality.

Contributor responsibility

Can you stand behind it?

Contributors are expected to do their best and understand the work they submit, including code assisted by an LLM.

Understand the patch

Project responsibility

Does it meet kernel standards?

AI assistance does not bypass review. Maintainability, correctness, and the project’s contribution process remain relevant.

Review the work

03 / What is being debated

Disclosure policy remains an open question

The report describes disagreement around declaring LLM use, but does not establish the final policy or how it will be applied.

“utterly stupid”

Torvalds, as quoted by XDA Developers
What remains unknown: XDA does not give the final wording or status of the disclosure policy, a measure of AI-assisted kernel submissions, or data on review time and acceptance outcomes. It reports concerns about AI-generated bug reports without counts or a comparison baseline. Calling this approach the “healthiest” is the article’s assessment, not a measured finding.

04 / The practical path

From tool choice to reviewed contribution

As maintainers handle AI-assisted work, the reported guidance leaves contributors responsible for producing a submission the project can review.

01

Choose tools

Contributors may use AI or work without it.

02

Understand

Know what the code does and be ready to stand behind it.

03

Submit

Send work through the project’s contribution process.

04

Review

Maintainers assess it against kernel requirements.

Why it matters: A ban could be difficult to apply consistently and restrict legitimate tools. Accepting generated code without careful review could burden maintainers. The reported approach keeps tool choice open while retaining responsibility for the submitted work.

05 / Key questions

What this position does—and does not—mean

Torvalds’ comments draw attention to the debate, but do not settle policy for every maintainer or open-source project.

Does Torvalds support AI-generated kernel code?

He supports using AI tools, and XDA reports he has used them in kernel-related work. Each patch still needs to meet Linux project standards.

Does he want contributors to disclose AI use?

XDA reports that he objected to a proposed approach around declaring LLM use. The report does not specify the policy’s final status or exact wording.

What does he expect from contributors?

As quoted by XDA, contributors should do their best and understand what they submit. The work remains their responsibility, whatever tools helped produce it.

Is this approval for unrestricted AI patches?

No. Support for AI as a tool is not approval for unreviewed code or submissions contributors do not understand. Kernel changes remain subject to review.

AI Use Meets Kernel Review Standards

The debate affects how a major open-source project handles a technology that can produce code quickly but does not remove the need for people to check it. A rule against AI use could be difficult to apply consistently and would prevent contributors from using tools Torvalds considers legitimate. On the other hand, accepting code simply because it was generated would risk burdening maintainers and the kernel with work that has not been carefully reviewed.

Torvalds’ reported position puts the emphasis on the quality and ownership of the patch, not on whether its author used an LLM. That distinction may inform discussions in other projects, although each community must decide its own review and disclosure policies. It also leaves room for contributors who prefer not to use AI, rather than making AI use a condition of participation.

There are costs to this approach. XDA notes that some contributors opposed to AI may be uncomfortable with its acceptance, while maintainers are already dealing with a volume of AI-generated bug reports. The article’s assessment that Torvalds’ approach is the “healthiest” is an opinion, not a measured finding about project outcomes.

Products Worth Considering

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite

LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDR True Black 400

LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz, 0.03ms, G-SYNC Compatible, FreeSync Premium Pro, HDR True Black 400

As an affiliate, we earn on qualifying purchases.

How Linux Reached the AI Debate

As coding agents and LLMs have improved, open-source projects have taken different approaches to AI-assisted contributions. According to XDA, some projects ban AI-generated code or commit messages, while others permit it under quality-control rules. The Linux debate described in the report centers partly on whether contributors must disclose AI use.

Torvalds is the original creator of the Linux kernel and a prominent figure in its development. His comments draw attention, but kernel changes still depend on the project’s contribution and review processes. The reported position is not a blanket permission for automated submissions: contributors are expected to understand what they submit, and the code must meet the project’s standards.

In discussing AI more broadly, Torvalds has described it as a tool to use appropriately and has praised its potential to make programming enjoyable for beginners. XDA’s account distinguishes that enthusiasm for experimentation from the requirements of kernel development, where maintainability and careful review remain important.

““utterly stupid””

— Linus Torvalds, as quoted by XDA Developers

Disclosure Rules Remain in Question

The report does not provide details about the final status or exact wording of the Linux project’s disclosure policy, nor does it describe a completed policy change across all kernel submissions. It also does not establish how maintainers will identify AI-assisted work when contributors do not disclose it.

It remains unclear how much AI-generated material is reaching kernel maintainers, how much time it adds to review, or whether AI use has changed the rate of accepted or rejected patches. XDA reports concerns about AI-generated bug reports, but gives no count, time period, or comparison baseline. Torvalds’ views also do not, by themselves, settle how individual maintainers should handle submissions.

Kernel Review Will Set the Practice

The immediate test is how maintainers apply existing review standards as contributors use AI tools and the project considers disclosure expectations. Contributors will still need to submit work that can be reviewed and meets kernel requirements; Torvalds’ comments do not announce an automatic acceptance path for AI-generated code.

Further clarity would come from a formal policy decision and examples of how maintainers handle AI-assisted patches and reports. Until then, the reported guidance is that tool choice is open, but responsibility for the submitted work remains with its contributor.

Key Questions

Does Linus Torvalds support AI-generated Linux kernel code?

He supports using AI tools, and XDA reports that he has used them in kernel-related work. That does not mean every generated patch is acceptable: submissions still need to meet Linux project standards.

Does Torvalds want contributors to disclose AI use?

XDA reports that Torvalds objected as maintainers moved to remove a requirement to declare LLM use. The article does not specify the final status or exact wording of the project’s disclosure policy.

What does Torvalds expect from contributors using AI?

As quoted by XDA, he expects contributors to do their best and understand what they are submitting. The principle is that the work must meet the project’s quality standards, whatever tools helped produce it.

Is Torvalds endorsing unrestricted AI-generated patches?

No. His reported support for AI as a tool is not an endorsement of unreviewed submissions or code contributors do not understand. Kernel changes remain subject to review.

Source: XDA Developers

You May Also Like

A Study Of Microsoft’s Early 2026 Rollout Of Claude Code And GitHub Copilot CLI

Microsoft is confirmed to launch Claude Code and GitHub Copilot CLI in early 2026, aiming to enhance developer tools with AI integration.

Fable and Mythos: How Anthropic Shipped Its Most Powerful Model to Everyone

Anthropic has launched Fable 5, its most capable model yet, with a safety architecture that allows broad access while maintaining security through fallback mechanisms.

Fable 5 Vs. GPT-5.6 Sol On An NP-Hard Problem: Does /Goal Help?

Fable 5 and GPT-5.6 Sol demonstrate differing capabilities on an NP-hard problem, with /goal command providing notable but limited assistance.

Muse Spark 1.1

Meta has launched Muse Spark 1.1, an AI model aimed at boosting creative applications, with initial evaluation reports highlighting its improved capabilities.