# Building a Security Framework for the Agentic Enterprise

DevFeed: [Building a Security Framework for the Agentic Enterprise](<https://devfeed.tech/articles/building-a-security-framework-for-the-agentic-enterprise-138077.md>)

Original publisher: [Read original article](<https://blogs.cisco.com/security/building-a-security-framework-for-the-agentic-enterprise>)

Author: Nancy Cam-Winget

Published: 2026-11-17T16:00:35Z

Content type: article

Language: en

Sources: [Cisco Blogs](<https://devfeed.tech/sources/cisco-blogs.md>)

Topics: [Security](<https://devfeed.tech/topics/security.md>), [shadow AI](<https://devfeed.tech/topics/shadow-ai.md>), [API Logging](<https://devfeed.tech/topics/api-logging.md>), [Cisco](<https://devfeed.tech/topics/cisco.md>)

Tags: [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agentic-enterprise](<https://devfeed.tech/tags/agentic-enterprise.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [apis](<https://devfeed.tech/tags/apis.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [architectures](<https://devfeed.tech/tags/architectures.md>), [authorization](<https://devfeed.tech/tags/authorization.md>), [cisco](<https://devfeed.tech/tags/cisco.md>), [connectivity](<https://devfeed.tech/tags/connectivity.md>), [context](<https://devfeed.tech/tags/context.md>), [controls](<https://devfeed.tech/tags/controls.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [data](<https://devfeed.tech/tags/data.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [framework](<https://devfeed.tech/tags/framework.md>), [identity-security](<https://devfeed.tech/tags/identity-security.md>), [security](<https://devfeed.tech/tags/security.md>), [security-framework](<https://devfeed.tech/tags/security-framework.md>)

## AI overview

Agent security requires identity, delegation, and policy context to persist across runtimes, identity services, APIs, MCP servers, and other agents so actions remain attributable and enforceable. The excerpt presents a vendor-neutral framework linking execution, identity, authorization, policy enforcement, connectivity, security and operational intelligence, and enterprise management through open interfaces. In its example, an agent presents a verifiable workload identity; an authorization service issues a short-lived, narrowly scoped token tied to the agent and, where relevant, its human or organizational delegator; enforcement evaluates context, and logging links the action to its owner, target, and policy decision. Unmanaged agents can be discovered, observed, and restricted while lacking verifiable identity. Adoption can be incremental and implementation details vary. The excerpt cites a survey in which 24% reported a recent incident involving compromised, excessive, or unauthorized access by an agent or other non-human identity, while 9% had fully integrated such activity into detection and response.

## Source excerpt

In this post, Cisco Fellow Nancy Cam-Winget describes why the industry needs more concrete, vendor-neutral security architectures for agents - and provides a concrete technical example.