When Ransomware Targets AI Models: Defending the AI/ML Recovery Chain

Ransomware crews have always followed the money. For years, that meant targets like banks and hospitals organizations with high-value assets and sensitive data who couldn’t afford downtime. Now, it means AI models too. A threat actor the Sysdig Threat Research Team (TRT) dubbed JADEPUFFER has already shown us where the future of ransomware is heading. In the span of a few weeks, JADEPUFFER went from poorly executed (albeit successful) database extortion with throwaway scripts to deploying a compiled ransomware binary built for one purpose: destroying AI and machine learning (ML) assets. That progression tells us what attackers now consider valuable. JADEPUFFER’s first campaign was scrappy, but it worked. The operator broke into a Langflow instance through CVE-2025-3248, an unauthenticated remote code execution (RCE) flaw that CISA added to its Known Exploited Vulnerabilities (KEV) catalog in May 2025. Langflow is a rich target because it holds LLM provider…