Overview

horizon3 platform for continuous, authorized autonomous autonomous-penetration-testing on live production systems. Chains misconfigurations, weak credentials and identity gaps into attack paths; provides fix guidance and re-verifies remediation. Can deploy honeypots during tests to detect live AI attackers.

Key Points

  • Marketed as production-safe with “zero downtime” across tests (vendor claim)
  • Data moat framed as hundreds of thousands of production autonomous tests
  • Complements runtime agent governance tools such as zenity (offensive validation vs intent control)

Sources