Executive Summary
JadePuffer is reported as the first AI-agent ransomware infection: it exploited a Langflow vulnerability and then encrypted MySQL databases and Alibaba Nacos.
What Happened
JadePufferと名付けられた史上初のAIエージェント型ランサムウェアの感染事例が報告され、Langflowの脆弱性を悪用して侵入した後、MySQLデータベースやAlibabaのNacosを暗号化攻撃の標的にしたとされている。[1] 記事では、中小企業がAIエージェントやLLMを業務に活用することで新たな攻撃対象となり、データ漏洩やAIモデル悪用などのリスクが高まると指摘している。[1]
Affected Products
- Langflow
- MySQL
- Alibaba Nacos
Vulnerability / Attack Technique
AI-agent-driven ransomware
Technical Mechanism
After gaining access through Langflow, the agent targeted databases and configuration services for encryption.
AI Security Category
AI-assisted attack
Part of the AI-Assisted Attacks topic.
Relevant Frameworks
RealGround Analysis
AI development tools are both the entry point and the attacker. RealGround analysis: Langflow and similar platforms must be patched and kept off the internet.
Recommended Controls
- Patch Langflow
- Keep backups of databases and configuration services offline
Related RealGround Service
Secure AI Agent Build · or start with the free AI Agent Security Scanner.
Sources
GIGAZINE（要約掲載：東京都中小企業サイバーセキュリティポータル）: https://www.cybersecurity.metro.tokyo.lg.jp/links/734/
Summary and analysis written by RealGround from the source above; last updated 2026-10-18.