Linus Torvalds Addresses AI-Generated Vulnerability Reports Overload in Linux Security System

Linus Torvalds has declared the Linux kernel's private security mailing list "almost entirely unmanageable" due to a flood of duplicate AI-generated vulnerability reports. In his weekly post to the Linux Kernel Mailing List (LKML), Torvalds outlined the challenges facing the security team as researchers increasingly employ automated tools to identify vulnerabilities, resulting in multiple reports of identical bugs.

The problem stems from researchers running similar AI tools against the same codebase independently and submitting findings separately to a private mailing list where previous reports aren't visible. This duplication has forced maintainers to spend excessive time triaging redundant reports and directing researchers to fixes that were already implemented weeks earlier.

"AI detected bugs are pretty much by definition not secret, and treating them on some private list is a waste of time for everybody involved," Torvalds wrote on LKML. This statement accompanied the release of Linux 7.1-rc4 and pointed to newly merged documentation formalizing how AI-assisted bug reports should be handled.

The scale of the issue is significant. According to Willy Tarreau, creator of HAProxy and a longtime Linux kernel stable maintainer, the kernel security mailing list received roughly two to three reports per week two years ago. Today, that volume has increased to five to ten reports daily, with most being legitimate findings but overwhelming the existing triage process due to duplication.

In response, Torvalds has directed developers to submit AI-detected vulnerabilities directly to relevant maintainers through a new public system rather than routing them through the private security list. The project's security bug documentation specifies that reports must be concise, formatted in plain text, and include a verified reproducer.

The Linux kernel project has formalized its broader stance on AI-assisted contributions, establishing project-wide policies that permit AI-generated code provided developers follow strict disclosure rules. Under this policy, AI agents cannot use the legally binding "Signed-off-by" tag, and contributors must use a new "Assisted-by" tag for transparency. Every line of AI-generated code, and any resulting bugs, remains the legal responsibility of the human who submits it.

Torvalds emphasized that researchers should go beyond simply filing raw findings. "If you actually want to add value, read the documentation, create a patch too, and add some real value on top of what the AI did," he wrote. "Don't be the drive-by 'send a random report with no real understanding' kind of person."

This approach aligns with fellow maintainer Greg Kroah-Hartman's "Clanker T1000" system—a Framework Desktop-powered bug-finding tool that follows a comprehensive workflow: discover the issue, write the fix, take responsibility for the patch, and submit it publicly.

The shift to a public reporting system reflects broader challenges in the software development ecosystem as AI tools become more prevalent. By moving to a transparent, public process, the Linux kernel project aims to reduce duplication, improve collaboration, and ensure that security researchers contribute meaningfully beyond what automated tools can provide.

{{IMAGE:2}}

The transition comes as the Linux kernel continues to serve as the foundation for countless systems worldwide, from embedded devices to supercomputers. Maintaining the security and stability of this critical software infrastructure requires efficient processes that can scale with the evolving landscape of vulnerability research.

For developers and security researchers working with the Linux kernel, the new documentation provides clear guidelines for responsible AI-assisted contributions. The project's stance balances technological advancement with accountability, recognizing the value of AI tools while emphasizing the irreplaceable role of human expertise in software development and security.