Radware (NASDAQ: RDWR) introduced a new product called Agent Trust Management on Sept. 29, 2026, designed to help organizations manage how AI agents interact with their applications.

The solution is intended to address a security gap that emerges as AI agents increasingly access business applications. According to Radware's H1 Global Threat Analysis Report, 77% of organizations are actively deploying or implementing AI agents, while only 17.2% report full visibility into AI agent activity within their environments.

Rather than applying broad allow-or-block controls, the product captures an agent's declared intent through direct conversation and continuously evaluates whether the agent's actions remain consistent with that intent. The system uses AI-based behavioral analysis, prompt profiling, and anomaly detection to assess agent activity over time.

Key capabilities include session-level visibility into agent identity and navigation, multi-signal identification of both compliant and non-compliant agents, and permission-based governance that allows organizations to restrict specific actions — such as login or checkout — while permitting others like browsing.

"AI agents are fundamentally changing how applications are accessed and used," said David Aviv, chief technology officer at Radware. "Organizations need to be able to embrace this shift while maintaining visibility and control over agent activity."

Radware Agent Trust Management is available as part of the company's Cloud Application Protection services, extending its existing bot management capabilities. The product targets use cases across e-commerce, financial services, and travel and hospitality, where AI agents may interact with sensitive data, transactions, and inventory systems.