Reflectiz has introduced a multi-agent penetration testing platform designed to identify, attack, and validate web vulnerabilities. The system utilizes specialized AI agents that operate based on pre-existing site models, which the company claims provides up to 10 times more coverage compared to conventional pentesting tools.
The platform functions by integrating an attacker’s perspective into the company’s existing live site models, which include data on pages, third-party scripts, and sensitive inputs. This approach allows the system to provide findings that include specific scripts involved, data reach, and current user exposure, enabling teams to move directly to remediation.
The new agentic pentesting is part of the company's Offensive Hub, which complements the existing Security Hub and Privacy Hub. The testing covers the full OWASP Top 10, with users able to set the depth of testing per flow, ranging from predefined checks to expert-level attack chains on critical assets.
Reflectiz co-founders Idan Cohen and Ysrael Gurt plan to host a live demonstration of the new platform on September 15 at 11 AM ET. The company provides continuous web exposure management services to enterprises across sectors including retail, finance, travel, insurance, healthcare, and gaming.
