Cisco acquires Astrix Security to tackle growing AI agent security challenges

Cisco Acquisition Targets Growing AI Agent Security Gap

Cisco Systems Inc. announced its acquisition of Israeli cybersecurity firm Astrix Security Ltd., marking a strategic move to address mounting concerns around AI agent security 1

1

. While Cisco did not disclose official terms, reports suggest the deal ranges between $250 million and $400 million, with Calcalist estimating approximately $400 million 2

2

. Founded in 2021, Astrix Security has raised $85 million in total funding from investors including Menlo Ventures and Anthropic 2

2

.

Securing Non-Human Identities Becomes Critical Priority

Peter Bailey, senior vice president and general manager of Cisco's Security Business Group, emphasized that Astrix Security specializes in protecting credentials that power modern systems, particularly API keys, OAuth tokens, and service accounts 1

1

. These are "the very credentials that AI agents are now using (and abusing) to gain access and execute work at scale," Bailey noted 2

2

. The platform developed by Astrix enables organizations to identify, manage, and secure AI agents operating in production environments, maintaining a real-time inventory of AI agents and non-human identities with contextual awareness of their activities 1

1

.

Identity Security Integration Expands Cisco's AI Defense Capabilities

Cisco plans to integrate Astrix's technology into its Cisco Identity Intelligence platform, enhancing visibility and context of identities within its security ecosystem 1

1

. The capabilities will extend to Cisco Secure Access and Duo Identity and Access Management offerings, allowing customers to authenticate and authorize agentic identities while responding to threats 1

1

2

2

. Key capabilities include AI agent discovery and governance, agentic access management, and threat detection and response for agents 2

2

.

Security Blind Spots Plague Enterprise AI Deployments

Cisco's AI Readiness Index reveals alarming statistics: only 24% of businesses can effectively control AI agents, while just 31% of enterprises consider themselves fully capable of securing their AI deployments 1

1

. Astrix co-founders Alon Jackson and Idan Gour highlighted that AI agents significantly outnumber human users, creating substantial security blind spots for many organizations 1

1

. The tools support agent lifecycle management and threat detection, essential for reducing attack surfaces and preventing compliance violations 1

1

.

Cisco Builds Comprehensive AI Security Portfolio

This acquisition follows Cisco's recent purchase of observability startup Galileo Technologies Inc., aimed at enhancing the trustworthiness and reliability of AI models 1

1

. At the RSAC 2026 Conference in March, Cisco introduced Zero Trust Access for AI agents, along with tools like Duo IAM for access management and DefenseClaw for scanning vulnerabilities 1

1

2

2

. Bailey emphasized that Cisco's differentiation lies in its "visibility across identity, network, application and infrastructure layers," enabling the company to understand not just what an agent is, but how it behaves 2

2

.