Human Security enhances AI agent security with Amazon Bedrock AgentCore integration

Cybersecurity firm Human Security Inc. has announced enhanced support for Amazon Bedrock AgentCore browser within its AgenticTrust service. This integration aims to ensure that artificial intelligence-powered agents built on Amazon Web Services Inc. infrastructure are cryptographically signed and secure, according to a report by Silicon Angle.

The collaboration addresses the emerging risks associated with agentic commerce, such as impersonation, misinformation, and fraud, which traditional bot management tools are ill-equipped to handle. Human Security's AgenticTrust service provides cryptographic verification and granular action-based controls, allowing businesses to define precisely what trusted AI agents can and cannot do. This enables enterprises to securely build, deploy, and manage AI agents, offering confidence in the authenticity and policy alignment of agent interactions. Customers can set specific rules, like allowing an agent to browse but not make purchases, thereby facilitating safer adoption of agentic commerce.

This partnership highlights the growing need for robust security frameworks as AI agents become more integrated into business operations. The integration of cryptographic verification and granular controls is crucial for establishing trust and mitigating risks in the evolving landscape of AI-driven automation.

Source: Silicon Angle