AI Risk Wire

Cases › security breach › LR-2026-0280

AI-assisted exploitation of PHP2Shell and Fragnesia flaws to breach 25 servers across four countries

Preliminary Why: 1 report: TechGig · subject to change · 1 independent report

Whose AI: Anthropic · Claude Codesecurity breach

Severity moderate: Twenty-five servers across four countries were breached, including a US government-affiliated organisation whose internal network and credentials were exposed.

Earliest source we hold 2026-10-29 · Captured by AI Risk Wire 2026-10-30 02:17 UTC

Attackers breached 25 servers in the US, Japan, Taiwan and South Korea, including systems tied to a US government-affiliated organisation, by exploiting PHP2Shell and a Linux flaw called Fragnesia.

Risk grouping: Not yet classified

No source we hold states: who caused it, intended or not, before or after release, the ai's part.

The story

1 · What it was doingAttackers used AI tools, including Claude Code and aiscan configured with DeepSeek and Kimi, to help breach 25 servers in the US, Japan, Taiwan and South Korea, including a US government-affiliated environment.
2 · What went wrongHackers exploited PHP2Shell and the Fragnesia Linux flaw to gain full control of vulnerable servers, breaching systems they were not authorised to access.
3 · What it reached or didTwenty-five servers across four countries were breached, and a US government-affiliated environment exposed its internal network and credentials.
4 · How it was caughtDream Research security researchers found evidence suggesting AI tools assisted, including about 40 custom scripts with consistent formatting.
5 · What was doneAdministrators were told to patch PHP installations, since updating WordPress plugins alone may not fix the vulnerability.

No source states: when it happened, when it was found, how many records were exposed, the direct loss, a count of the other parties affected, any court case, the cause, what set it off.

Others like this

Indonesian-speaking operator uses Claude Code to compromise Indonesian and Bangladeshi government domains 2026-10-26

CyberXero AI-augmented initial access campaign using Claude Code and PentAGI agents 2026-10-12

Agentic AI intrusions by four threat actors (CrowdStrike 2026 Global Threat Report) 2026-10-28

Jailbroken Claude used to breach Mexican government networks and steal taxpayer and voter records 2026-03-02

All cases of this kind ›

Show every field, with its source and grade

The AI system

Model
Claude Code
C one secondary source states it
Developer
Anthropic
C one secondary source states it
Software around the model
Claude Code and aiscan
C one secondary source states it

What happened

Earliest source we hold
2026-10-29
A the date the source itself carries
Themes
AI assisted cyberattack, server breach, government linked systems
C one secondary source states it
Who caused it
an outside attacker
C one secondary source states it
What went wrong
security breach
C one secondary source states it

Impact

Severity
moderate
D inferred by us from the described behaviour, reason shown
Kind of harm
others' security, operations
C one secondary source states it
Who was affected
25 servers across the United States, Japan, Taiwan and South Korea, including systems connected to a U.S. government-affiliated organisation
C one secondary source states it
Kinds of data exposed
credentials
C one secondary source states it
Why this severity
Twenty-five servers across four countries were breached, including a US government-affiliated organisation whose internal network and credentials were exposed.
D inferred by us from the described behaviour, reason shown

LR-2026-0280 · version 2 · published 2026-10-30 after automatic checks; not yet checked field by field by a person · changes to this case · JSON