Greg Kroah-Hartman on the LLM bug deluge: “don’t panic”

Translated by a model; the original is in Russian.

· Security
At Kernel Recipes 2026, Greg Kroah-Hartman explained how the Linux kernel's security team is coping with the flood of vulnerability reports unleashed by large language models.

cover.svg

Large language models have turned finding security bugs into something almost anyone can do. The consequence is predictable: nearly every free-software project — the Linux kernel included — is now swamped with security bug reports. At the 2026 edition of Kernel Recipes, Greg Kroah-Hartman took the stage to describe how the kernel's security team is handling the surge.

His central message was strikingly calm: “don't panic.”

There is a clear logic behind that short phrase. LLM-based tools have lowered the barrier to bug hunting so far that reports now arrive in unprecedented volume — and not all of them are equally useful. For maintainers, this is less a new threat than a new burden: separating signal from noise at a scale the old processes were never designed for.

That is why “don't panic” reads not as a brush-off but as a working stance. The kernel's security team has long dealt with a steady stream of reports, and the arrival of another source — however powerful — is no reason to overhaul the fundamental approach. Keeping a steady hand and relying on established procedures matters more than treating every machine-generated warning as a crisis.


Source: LWN.net

Comments

Sign in, to comment.

  • No comments yet.