Anthropic says its AI model built a working root exploit for critical Linux kernel flaw CVE-2026-98432
CRITICAL · VULNERABILITYCVE-2026-98432redteam/daily · 6 sources
- Severity
- Critical
- Highest CVSS
- 9.8
- Exploited
- Not known
- Sources
- 6
What happened
Anthropic reported that an unreleased AI model wrote a working attack against a serious bug in the Linux operating system's core code. The bug could let a remote attacker take full control of a server without logging in, though Anthropic says this was a lab test. Fixes exist upstream, and administrators should install their distribution's kernel update and restart their systems.
Like I’m new · Written by AI from 6 sources · Wrong?
What to do
- Check whether you use the affected product, and install the vendor’s fix when it is available. Open your watchlist
General advice for this kind of story
Vulnerabilities
Full story timeline · 7 updates
- Thu 15 OctoberCVE-2026-98432 published, CVSS 9.8NVD
- Thu 29 OctoberClaude AI Exploits Linux Kernel Flaw to Execute Root Commands Without User InteractionCyber Security News
- Linux Kernel TCP Flaw With 9.8 CVSS Score Affects Ubuntu, Red Hat and Other DistributionsThe Hacker News
- Fri 30 OctoberCritical Linux Kernel Vulnerability Lets Remote Attackers Crash Ubuntu SystemsGBHackers
- Anthropic AI model finds Linux kernel bug that leads to remote root accessCSO Online
- Linux security advisory (AV26-1108)Canadian Centre for Cyber Security
- Sat 31 OctoberCVE-2026-98432 - Linux kernel IPv4 TCP use-after-free IssueAWS Security Bulletins
Who reported · 6 outlets
Cyber Security NewsThe Hacker NewsGBHackersCSO OnlineCanadian Centre for Cyber SecurityAWS Security Bulletins
Who and what is in it
Companies