Cisco announced a broad set of security products at RSAC 2026 Monday aimed at securing the growing use of AI agents in enterprise environments. The announcements span identity management, pre-deployment testing, open-source tooling, and SOC automation.

The centerpiece is Zero Trust Access for AI agents, which extends Cisco’s existing access control model to cover agentic workloads. New Duo IAM capabilities let organizations register agents, map them to accountable human owners, and enforce fine-grained, task-specific permissions. All agent tool traffic is routed through an MCP gateway to eliminate blind spots. Cisco Identity Intelligence handles discovery of existing agentic and non-human identities across the environment.

“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. “Security teams are the key to unlocking this opportunity by making the agentic workforce safe enough to trust.”

A recent Cisco survey found 85% of major enterprise customers are experimenting with AI agents, but only 5% have moved them into production. The access control gap is a primary reason.

On the testing side, Cisco is launching AI Defense: Explorer Edition, a self-service version of the AI Defense Validation engine used by Global 2000 customers. It supports multi-turn adversarial testing, prompt injection and jailbreak testing, and CI/CD integration for GitHub Actions, GitLab, and Jenkins. A companion Agent Runtime SDK embeds policy enforcement directly into agent workflows at build time, with support for AWS Bedrock AgentCore, Google Vertex Agent Builder, Azure AI Foundry, and LangChain.

Cisco is also releasing DefenseClaw, an open-source secure agent framework that bundles Skills Scanner, MCP Scanner, AI Bill of Materials, and CodeGuard into a single tool. DefenseClaw will integrate with NVIDIA’s OpenShell for runtime security.

On the SOC side, Splunk is shipping Detection Studio for managing the full detection engineering lifecycle, Federated Search for correlating data across environments, and Exposure Analytics for continuous asset visibility. Several specialized AI agents, including a Triage Agent, Malware Threat Reversing Agent, and Detection Builder Agent, are in various stages of release through June 2026.

Detection Studio and the Malware Threat Reversing Agent are generally available now. Exposure Analytics, the SOP Agent, and Federated Search are expected in April and May.