Today, Workday introduced Agent Passport with Cisco as its launch partner. What this unveils is a new standard for the industry—a model where the builders of agentic platforms can deliver Agent Security as a native capability to their customers and developers. 

Why agent security matters

Agents are changing what enterprise security must do. They do not sit behind a firewall and serve up pages — they act on sensitive enterprise workflows, run payrolls, analyze financial data, manage employee records — all at a speed humans cannot supervise transaction by transaction. The security model must move with them.

For decades, enterprises added security in layers around their workloads. While that model works well at the infrastructure layer, it needs to be reimagined for millions of agents. Security that sits outside the agent is stuck reacting to decisions already after they’ve been executed. With Cisco AI Defense, we close this gap. For four decades, Cisco has secured the networks, identities, applications, and data the enterprise runs on, always by putting protection directly in the path of the workload, never as an afterthought. We are bringing the same discipline to agents: validating them against the attack classes that matter before they ship and protecting them with runtime guardrails after they’re deployed.

Cisco AI Defense created an entirely new category of AI security: enterprise agentic security delivered from inside the platforms agents run on, aligned to their identities, policies and data boundaries, and surfaced to customers and builders in the same workflows they already use to ship and govern agents.

Enabling the builders of agentic platforms

At the Cisco AI Defense team, we view this as our most strategic priority: empowering the builders of agentic platforms to embed enterprise-grade security natively into their offerings

Every company building an agentic platform is being asked the same question by their customers. Whether the workload is for HR and finance, developer tooling, customer service, or another industry-specific domain, they all are asking, “How do I know the agents on your platform are safe?”

Today, that question is hard to answer. With Cisco AI Defense, we want to make it easy. Our goal is to give the builders of agentic platforms AI security they can embed natively. From the moment a developer or ISV ships an agent on their platform, it can be tested against the most serious risks and then carry a verifiable, signed record of what evaluations it has passed. And as a result, everyone wins. Platform builders get a trust layer they did not have to invent. Developers get a clear signal that helps their agents reach production faster. Customers get a consistent answer to the safety question, no matter which agent they enable.

Cisco and Workday: Partners in defining enterprise agentic security

Workday Agent Passport, announced today at Workday DevCon 2026, brings Cisco’s enterprise agentic security to life in production. We are honored to be the launch partner. Together with the Workday team, we are giving Workday customers and developers a direct path to Cisco AI Defense evaluations from inside the Workday developer platform. In practice, that means Workday developers can put their agents through a Cisco AI Defense evaluation before they go live, confirming the agent stays on task, protects sensitive information, and avoids producing unsafe results. In the future, there may be opportunities to explore protections that extend across more of an agent’s lifecycle, from pre-deployment through runtime.

What distinguishes this partnership is the depth and discipline behind it. From day one, the Workday team has set a high bar: grounded in open industry standards, principled in giving customers their own choice of trusted providers and deliberate about reducing friction for the partners and developers building on the platform.

To the builders of agentic platforms

Trust is the gating factor on the agentic AI transformation. Customers will deploy agents at the scale this moment deserves only when they can answer who tested the agent, what was tested and how. Platform-level agent security is how we make that answer an easy one for customers, developers, and the builders of the platforms where this transformation will run.

Our announcement today with Workday is what the first version looks like. If you are building an agentic platform, this is the conversation we want to have with you next.

Learn more

• Cisco AI Defense: cisco.com/go/ai-defense
• Workday Agent Passport: workday.com/agent-passport
• Workday’s launch announcement 
• Building an agentic platform? Reach out: ai-defense-partners@cisco.com