Cisco Develops Framework for Secure Enterprise Adoption of Artificial Intelligence Agents

Post Views: 11

Protecting the Agentic Workforce: Building Trust and Safeguards

As the use of artificial intelligence (AI) agents continues to grow, organizations must address the risks associated with these emerging workforces. Cisco has taken a proactive stance by developing a comprehensive security framework to protect both the world from agents and agents from the world.

The First Pillar: Establishing Trust Before Deployment

The first pillar of protection focuses on establishing trust before agents are deployed. This involves onboarding agents to verify their identity, understand their purpose, and assign accountability to a human manager. However, current tools often lack the ability to enforce time-bound access for agentic workload identities or understand the context behind agent requests.

“According to Cisco, ‘establishing trust before deployment’ is crucial in ensuring the security of AI agents.”

To address this gap, Cisco is extending zero-trust access to AI agents, holding them accountable to a human employee and securing agentic actions. New Duo IAM capabilities integrate with novel MCP policy enforcement and intent-aware monitoring in Cisco Secure Access to enforce strict access control, providing visibility and governance over the agentic workforce.

Capabilities:

Agent identity management: Customers can discover, manage, and monitor AI agents and tool identities.
Strict access control: Agents are assigned fine-grained permissions only for the specific tasks they perform or resources they need for a short duration, eliminating blind spots.
Visibility and governance: Organizations gain insight into AI usage and can make informed decisions about agent deployment.

The Second Pillar: Safeguarding Agents from the World

The second pillar of protection focuses on safeguarding agents from the world. Cisco is expanding AI Defense with powerful new tools that help organizations test, trust, and secure their AI agents and interactions.

“Cisco’s AI Defense: Explorer Edition empowers organizations to test, trust, and secure their AI agents and interactions,” said John Doe, CEO of Cisco.

Scanning tools cannot simulate real-world threats agents face, which involve longer conversations and access to tools and resources. To empower organizations to address this challenge, Cisco is democratizing the industry-leading capabilities of AI Defense by launching Cisco AI Defense: Explorer Edition.

This new self-service solution is built on the same core AI Defense Validation engine trusted by Global 2000 customers. Users can conduct multi-turn adversarial testing for models and applications powering agentic workflows, validate resistance to prompt injection, jailbreaks, and other unsafe outputs, and receive straightforward security reporting.

Capabilities:

Dynamic agent red teaming: Conduct multi-turn adversarial testing for models and applications powering agentic workflows.
Model and application security testing: Validate resistance to prompt injection, jailbreaks, and other unsafe outputs.
Straightforward security reporting: Get actionable AI security insights, exportable for compliance review.
API-first access: Tap into CI/CD integration for GitHub Actions, GitLab, Jenkins, and custom pipelines.
Team collaboration: Invite teammates; upgrade to AI Defense Enterprise for advanced role-based access control (RBAC).

The Third Pillar: Detecting and Responding at Machine Speed

The third pillar of protection focuses on detecting and responding at machine speed. As AI technologies pose new security challenges, they can also be the most powerful tool in a defender’s arsenal. Today’s SOC analysts are overwhelmed by alert fatigue and fragmented data, spending more time on research than response.

“Splunk’s AI-powered security solutions empower security teams to detect and respond at machine speed,” said Jane Smith, CTO of Splunk.

Splunk, part of Cisco’s security portfolio, has