EventCybersecurityJuly 2, 2026
AI agent runs first fully automated ransomware attack
Sysdig reports an AI agent, dubbed JADEPUFFER, exploited CVE-2025-3248 in Langflow to break into servers, steal credentials, and encrypt a MySQL database. The flaw was patched in Langflow 1.3.0.
1 source
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- AI trading bot for Hyperliquid uses LLM models
- Robert Scoble's robot drives him to Starcloud
- AI/ML penetration testing roadmap covers prompt injection and LLM attacks
- Sentinel framework repurposes Android devices as AI monitoring nodes
- User uses Claude to draft letter disputing $1,000 cancellation fee