Pentera Expands Security Validation to AI-Native Web Applications

The new capability allows organizations to run continuous testing within defined production guardrails. Unlike traditional signature-based scanners, these AI agents perform reconnaissance, navigate complex application logic, and chain multiple vulnerabilities together to demonstrate actual business impact. The system leverages a decade of offensive research and real-world attack data to adapt payloads in real time, identifying flaws in authorization and business logic that often evade standard automated tools.

CEO Amitai Ratzon noted that by consolidating infrastructure and application testing into a single platform, the company intends to simplify offensive security workflows. The technology supports authenticated testing environments, including SSO, MFA, and OAuth, and generates actionable evidence intended for developer remediation. While currently in beta for select clients, the feature is scheduled for general release in the fourth quarter of 2026. Pentera plans to showcase the live platform at the Black Hat USA 2026 conference in early August.