Cisco Systems Inc. (NASDAQ: CSCO) announced new security solutions designed for AI agents that perform actions rather than just answer questions. The company unveiled these capabilities at RSA Conference 2026 in San Francisco.

The networking equipment maker is extending its Zero Trust Access technology to AI agents through new features in Cisco Identity Intelligence and Duo identity management systems. The company will provide agent discovery capabilities, identity and access management for AI agents, and policy enforcement through its Secure Access security service edge platform.

Cisco also launched AI Defense: Explorer Edition, a self-service tool that allows developers to test AI models and applications against potential attacks before deployment. The platform includes dynamic agent testing, security validation, and integration capabilities for continuous development pipelines.

The company introduced DefenseClaw, an open source framework for securing AI agents that automates security processes and inventory management. Cisco plans to integrate this with NVIDIA's OpenShell sandbox environment.

According to a Cisco survey of enterprise customers, 85% reported experimenting with AI agents, but only 5% have moved the technology into production environments.

"AI agents aren't just making existing work faster; they're a new workforce of co-workers that dramatically expand what organizations can accomplish," said Jeetu Patel, president and chief product officer at Cisco.

For security operations centers, Cisco's Splunk division added new AI-powered capabilities including Exposure Analytics for asset inventory and risk scoring, Detection Studio for managing security detection workflows, and specialized AI agents for various security tasks including malware analysis and threat response.

Some features are currently available while others are scheduled for release between April and June 2026, with additional capabilities targeting June 2026 for testing phases.