JadePufferThreat actor
JADEPUFFER is an agentic threat actor that executed a fully autonomous ransomware operation, leveraging a Large Language Model to automate the entire attack chain from initial access to data destruction. It exploited CVE-2025-3248 against an exposed Langflow instance for initial access, then compromised MinIO using default credentials and manipulated MySQL for privilege escalation.From MISP Galaxy
Last reported 15 days ago, by Infosecurity Magazine7 reports
ATT&CK techniques
4 techniques · each with the sentence it comes from- T1684Social Engineering
“…begins with initial access, with AI models used to discover vulnerabilities in source code, binaries and AI serving systems, while social engineering campaigns such as phishing can be customized at scale.”
- T1053.005Scheduled Task
“The actor uses a new RedFlick delivery technique to install the CosmicPulse backdoor through scheduled tasks, supporting espionage activity against governments, NGOs, think tanks, and organizations connected to Ukraine.”
- T1204.003Malicious Image
“Processing a malicious image or PDF can allow arbitrary code execution.”
- T1505.003Web Shell
“Attackers have exploited the flaws to gain remote access, deploy web shells and tunneling malware, steal credentials, and move from exposed appliances into internal networks.”
Reports per month
last 12 monthsOctober so far (hatched). Counted by report, not by story.
What the reports say
7 reports, newest first · the exact sentence from each-
Q3 2026 Sets New Record for Ransomware Attacks“In July, researchers identified the JadePuffer campaign, believed to be the world’s first ransomware attack completely driven by AI.”
-
5th October – Threat Intelligence Report“Researchers outlined how JadePuffer, an AI-enabled threat actor tracked as Storm-3168, used compromised Azure service principals to automate cloud reconnaissance and destructive actions.”
-
Microsoft: AI Cuts Post-Compromise Attack Time to Minutes“For example, it highlighted how threat actors are starting to transition to fully autonomous AI attacks, such as the JadePuffer campaign identified in July.”
-
JadePuffer crims hijacked Azure identities and used them to blow up cloud resources“JadePuffer crims hijacked Azure identities and used them to blow up cloud resources”
-
JadePuffer AI Actor Compromises Azure Tenant in Destructive Cloud Attack“JadePuffer AI Actor Compromises Azure Tenant in Destructive Cloud Attack”
-
Cloud and SaaS Environments Now Top Targets for Attackers“In July, the world’s first fully AI-generated ransomware campaign, dubbed JadePuffer, was highlighted by security researchers.”
-
JadePuffer Returns With Ransomware Designed to Wipe AI Models“JadePuffer Returns With Ransomware Designed to Wipe AI Models”
Thanks — we review every report.