In this interview from the AWS Marketplace Series 2026, Shahar Peled, co-founder and chief executive officer of Terra Security, joins Anna Sarnek, vice president of business and strategy at Terra Security, to talk with theCUBE Research's Christophe Bertrand about the shift from periodic penetration testing to AI-powered continuous offensive security.
Peled explains why the convergence of AI-accelerated development and AI-equipped adversaries has made traditional, once-a-year testing fundamentally inadequate. Terra Security responds with a swarm of hundreds of AI agents trained to execute the full offensive security cycle — from business context collection and attack surface analysis to vulnerability validation, fix generation and continuous retesting — ensuring organizations remain secure beyond any single point in time.
The conversation also explores Terra's landmark position as the first company to earn AWS Security Competency certification under the continuous autonomous validation category for AppSec, a use case developed in close collaboration with Amazon. Sarnek details how building 100% on AWS Bedrock gives enterprise customers in healthcare, financial services and high-tech the confidence to adopt Terra through AWS Marketplace, where streamlined procurement and trusted ecosystem validation help shorten sales cycles.
A real-world example from a major financial organization illustrates the stakes: Terra's agents detected a chain of seemingly minor vulnerabilities that together constituted a remote code execution (RCE) flaw — a finding that would have been invisible to manual or point-in-time testing, but fully exploitable by today's AI-powered adversaries.
From bridging the gap between security and IT teams to translating technical risk into board-level business impact, Peled and Sarnek outline why continuous autonomous validation is rapidly becoming a baseline requirement for enterprise cyber resilience.
