CrowdStrike Details Four Threat Actors Using Agentic AI in Intrusions
Summary
CrowdStrike reports that four threat actors used AI agents, including Claude Code, in intrusion campaigns between May and August 2026. The cases suggest agentic AI may lower barriers to offensive operations; the company assesses that activity will likely increase.
Key points
- CrowdStrike documented four previously unreported threat actors using AI agents to make operational decisions and execute commands during campaigns between May and August 2026.
- One actor compromised at least 15 Argentina-based healthcare entities and exfiltrated credentials and medical records from at least one victim.
- A Chinese-speaking operator had more than 20 named global targets and used two pentesting harnesses that gave Claude Code different levels of autonomy.
- An Indonesian-speaking actor used 819 skills, mostly from public GitHub repositories, to guide Claude Code intrusions, sometimes intervening when safety guardrails or other barriers impeded progress.
- A Russian-speaking group reportedly exfiltrated corporate client data from a U.S. CRM vendor and compromised private-sector SaaS and data center entities in several countries.
- CrowdStrike assesses with moderate confidence that agentic AI is commoditizing baseline offensive operations; it expects AI-supported intrusions to increase in volume and sophistication.
Article Details
- Publisher
- CrowdStrike
- Report Period
- 2026-05 to 2026-08
- Scope
- Four previously unreported threat actors and their agentic AI-supported campaigns.
- Sample Size
- Four threat actors; CrowdStrike documented several incidents.
- Key Statistics
- Threat Actor 1 compromised at least 15 Argentina-based healthcare entities.
- Threat Actor 2 had more than 20 named global targets.
- Threat Actor 3 used 819 skills, overwhelmingly sourced from public GitHub repositories.
- The report says agentic AI can autonomously achieve objectives, often within an hour.
Products
Tools
claude-red1actor that uses 819 skills (overwhelmingly sourced from public GitHub repositories, primarily claude-red1) without additional orchestration to enable Claude Code to conduct intrusion campaigns.PentestGPTThis framework is derived from the open-source PentestGPT project, but the threat actor has rebuilt it with additional MCP servers, Python-based exploitation scripts, and skills.
Countries
ArgentinaA Spanish/English-speaking operator that has compromised at least 15 Argentina-based healthcare entities and exfiltrated credentials and medical records from at least one victim.BrazilThey compromised government entities of two countries and targeted private sector entities in Brazil and the U.S., including entities in biotech, healthcare, food and beverage, retail, technology, and travel.Kyrgyzstanmanagement vendor and compromised several private-sector SaaS and data center entities based in Russia, Kyrgyzstan, and the U.S., in addition to one government domain of another country.Russiamanagement vendor and compromised several private-sector SaaS and data center entities based in Russia, Kyrgyzstan, and the U.S., in addition to one government domain of another country.
Industries
Customer relationship managementThey exfiltrated corporate client data from a U.S.-based customer relationship management vendor and compromised several private-sector SaaS and data center entities based in Russia, Kyrgyzstan, and the U.S., inData centera U.S.-based customer relationship management vendor and compromised several private-sector SaaS and data center entities based in Russia, Kyrgyzstan, and the U.S., in addition to one government domain of anotherFood and BeverageThey compromised government entities of two countries and targeted private sector entities in Brazil and the U.S., including entities in biotech, healthcare, food and beverage, retail, technology, and travel.HealthcareA Spanish/English-speaking operator that has compromised at least 15 Argentina-based healthcare entities and exfiltrated credentials and medical records from at least one victim.RetailThey compromised government entities of two countries and targeted private sector entities in Brazil and the U.S., including entities in biotech, healthcare, food and beverage, retail, technology, and travel.TechnologyThey compromised government entities of two countries and targeted private sector entities in Brazil and the U.S., including entities in biotech, healthcare, food and beverage, retail, technology, and travel.TravelThey compromised government entities of two countries and targeted private sector entities in Brazil and the U.S., including entities in biotech, healthcare, food and beverage, retail, technology, and travel.