OPEN LINK ↗
// 1h agoSECURITY INCIDENT
Claude Turns Linux Bug Into Remote Root Exploit
Anthropic says an unreleased Claude research model turned Linux CVE-2026-98432 into a zero-click exploit capable of running commands as root on servers with an MPTCP listener. Upstream and Debian fixes are available, while Ubuntu and RHEL still list affected or pending packages.
// ANALYSIS
This is a meaningful escalation in AI-assisted vulnerability research: the model did not merely identify a bug, but chained memory corruption, KASLR bypass, and control-flow hijacking into a working exploit.
- –The attack requires an exposed multipath TCP listener, narrowing the target set but eliminating authentication and user interaction
- –Anthropic reports roughly 48 minutes to leak a kernel pointer, followed by five minutes to reach command execution
- –Debian shipped fixes quickly, highlighting the operational risk of uneven downstream patch timelines
- –Ubuntu and RHEL administrators should inventory MPTCP listeners and prioritize kernel updates or mitigations
- –The disclosure raises the bar for defensive testing as frontier models become capable of exploit development, not just code analysis
// TAGS
claudellmsecurityresearchreasoning
DISCOVERED
1h ago
2026-10-30
PUBLISHED
1h ago
2026-10-30
RELEVANCE
9/ 10
AUTHOR
IntCyberDigest