AI Wiki · Mediocre Security
IncidentIncidents

PHP2Shell intrusions

An attacker rooted 25 servers, three in a U.S. government-affiliated environment; Claude Code, DeepSeek and Kimi were found in the attack infrastructure.

What happened

an attacker chained a custom PHP2Shell exploit with a Linux kernel flaw to root 25 servers, including three in a U.S. government-affiliated environment, after scanning about 19,000 exposed firewalls. Dream found Claude Code plus DeepSeek and Kimi in the attack infrastructure, but there is no direct evidence AI wrote the exploit.

Check Point said on Nov 13 2026 that attackers entered its Mobile Access portal through PHP2Shell, then used a Gaia OS zero-day (CVE-2026-123039, a root escalation through an unauthenticated local socket) to take over about two dozen customers’ gateways, install backdoors, pivot to VMware ESXi and deploy wiper malware. CISA added it to KEV the same day; no report says AI was used for this step.

German host Mittwald says the attacker in the Mittwald MCP server breach first tested its MCP flaw during the October scramble to patch PHP2Shell.

Timeline

Every dated event in the wiki that involves PHP2Shell intrusions, newest first.

Connections

PHP2Shell intrusions and everything one link away. Drag to move, scroll to zoom, click a node to open it.Open in the full graph →

What links here

Concept AI and CVEs: latent flaws, now exploitable

Company Calif · Check Point

Model family Claude · Kimi

Product Claude Code

AI lab DeepSeek · Moonshot AI (Kimi)

Incident Mittwald MCP server breach

Sources

  1. The Record, “Limited reports of PHP bug exploitation emerge”, Oct 29 2026
  2. Cybersecurity Dive, “Palo Alto Networks firewalls compromised in chained attacks using PHP2Shell”, Oct 30 2026
  3. BleepingComputer, “Check Point warns of zero-day flaw exploited for root access”, Nov 13 2026