Claude wrote a Linux root exploit without a password, and Ubuntu still has no fix

Anthropic's Claude AI model generated a working remote root exploit for a Linux kernel vulnerability tracked as CVE-2026-98432, allowing passwordless privilege escalation on affected systems. Red Hat has released a patch for the flaw, but Ubuntu has not yet shipped a fix, leaving its users exposed. The event matters because it demonstrates AI's ability to weaponize kernel vulnerabilities into functional exploits, shortening the window between disclosure and active attack. Operators running Ubuntu nodes should watch for an official patch and consider mitigations such as restricting remote access in the interim.