Armadin and TENEX.ai Run Controlled AI Cyberattack Test at Global Institution

Armadin and TENEX.ai said in a press release that they ran a three day controlled AI cyberattack test for a leading global institution. Armadin's autonomous attacker operated across the institution's external perimeter, internal network, and web applications, while TENEX.ai handled detection, triage, and response.

Armadin said its swarm generated 17 million offensive actions during the exercise. The system launched 1,300 attacks using 26,000 agents, targeted more than 25,000 services, and produced 238 security findings, including 38 validated attack paths.

TENEX.ai said its security operations platform triaged all 101,169 alerts created across the institution's security stack. Its agents traced activity across 231 billion raw events and reconstructed the 38 attack paths, while human analysts made investigation and escalation decisions.

The companies said a manual investigation of the same scope would have required about 2,400 analyst hours. TENEX.ai also produced evidence based determinations for all 238 findings, including affected endpoints, available telemetry, existing detections, and any additional detection logic required.