AI Agent Swarm Compromises Company in 14 Minutes
David Kennedy reports an AI agent swarm bypassed security and deployed ransomware in 14 minutes, far exceeding previous AI-driven breach speeds.
David Kennedy, founder of cybersecurity firm TrustedSec and a former NSA hacker, reported that an AI agent swarm compromised a client company in 14 minutes. The attack utilized valid credentials and more than 30 different types of AI agents to probe infrastructure simultaneously, bypassing two-factor authentication to exfiltrate data and deploy ransomware.
The victim company failed to detect the intrusion for approximately four and a half hours, only discovering the breach after the encrypted system crashed. Kennedy stated that the speed of this attack significantly exceeds previous benchmarks, including a September report by Palo Alto Networks describing a 10-hour AI-driven breach.
This incident follows a November warning from the Federal Bureau of Investigation regarding the Chinese-built ARTEX framework, which allows privileged access in under three hours. Kennedy noted that the overall message is that these attacks are happening quicker than organizations can typically respond.