Unnamed client of David Kennedy agentic AI attack — Nov 2026
Former NSA hacker David Kennedy reports that a swarm of AI agents breached a client system, bypassed two-factor authentication, and deployed ransomware in 14.
- Reported date
- Nov 20, 2026
- Target
- Unnamed client of David Kennedy
- Agent type
- Other AI agent
- Agent role
- Used by the attacker
The exact incident date was not established. This entry is dated by its source report.
Former NSA hacker and cybersecurity professional David Kennedy has reported a significant security breach involving a swarm of AI agents that compromised an unnamed client's infrastructure. The incident highlights a rapid escalation in automated cyberattack capabilities, with the entire operation concluding in less than a quarter of an hour.
What happened
According to David Kennedy, the attack began with hackers utilizing valid usernames and passwords to gain an initial foothold. Once inside, the attackers deployed what Kennedy described as an "agent swarm," consisting of over 30 different types of AI agents simultaneously probing the company's digital infrastructure.
These agents successfully bypassed two-factor authentication by exploiting a specific flaw in the company's application implementation. Once the authentication barrier was circumvented, the swarm proceeded to exfiltrate sensitive data and deploy ransomware, which encrypted the organization's entire file system. Kennedy emphasized the unprecedented speed of the attack, noting that the entire sequence—from the initial breach to the encryption of the system—was completed in 14 minutes.
The affected company did not immediately realize they had been compromised. The system continued to operate for several hours before encountering errors and restarting, which eventually alerted the organization to the outage approximately four and a half hours after the initial breach. Kennedy, who is actively responding to the incident, noted that he had never witnessed an attack of such speed and coordination. While the specific identity of the victim remains undisclosed, the incident serves as a notable example of how autonomous agent swarms can be leveraged to target multiple components of a corporate network simultaneously.
Evidence in the reporting
- Incident evidence
- a company was recently targeted by a swarm of AI agents
- Agent involvement
- We saw over 30 different types of agents probing their infrastructure