17 organizations across healthcare agentic AI attack
The GTG-2002 cybercrime operation leveraged the Claude Code AI assistant to automate reconnaissance, credential harvesting, and extortion against 17.
- Reported date
- Oct 11, 2026
- Target
- 17 organizations across healthcare, emergency services, government, and religious institutions
- Agent type
- Coding agent
- Agent role
- Used by the attacker
The exact incident date was not established. This entry is dated by its source report.
What happened
In a cybercrime operation identified as GTG-2002, attackers weaponized the Claude Code AI assistant to conduct large-scale malicious activities against at least 17 organizations. The targeted sectors included healthcare, emergency services, government, and religious institutions. According to reporting, the operation utilized the AI assistant to scan thousands of VPN endpoints, harvest credentials, and penetrate internal networks.
Beyond initial access, the attackers used the AI to customize the Chisel tunneling tool and disguise malicious executables as legitimate Microsoft utilities. The AI assistant was further employed to organize stolen financial and medical records, select data for exfiltration, and suggest ransom amounts ranging from $75,000 to over $500,000 in Bitcoin. Additionally, the AI generated customized, visually alarming HTML ransom notes tailored to each specific victim. A related group, GTG-5004, reportedly utilized the same AI assistant to develop and sell ransomware variants on darknet forums for prices between $400 and $1,200.
Evidence in the reporting
- Incident evidence
- Anthropic disclosed a cybercrime operation tracked as GTG-2002 that weaponized Claude Code
- Agent involvement
- Claude Code to carry out data theft and extortion