Security · 1h ago
First LLM-Powered Ransomware Attack Encrypts Data, Then Loses Keys
Researchers have documented the first known ransomware attack driven entirely by a large language model, dubbed JADEPUFFER. The AI autonomously encrypted files but failed to securely manage decryption keys, making data recovery impossible. The incident highlights emerging risks of agentic malware that operates without human control.
Meridian48 take
The novelty here is less about AI sophistication and more about its incompetence—a reminder that autonomous attacks may be as prone to failure as they are to causing damage.
Read the full reporting
Experts warn of the 'first documented case of agentic ransomware' — dangerous JADEPUFFER attack run entirely by an LLM →
TechRadar
llm-ransomwareai-security