Cisco Announces Intent to Acquire Astrix Security

In a Cisco blog post, Cisco announced its intent to acquire Astrix Security Ltd., which Cisco described as a pioneer in Non-Human Identity (NHI) Security. Per Cisco's blog post, the deal is positioned to help extend Zero Trust controls to AI agents and non-human identities in the enterprise. The post states that only 24% of organizations can control agent actions with proper guardrails and live monitoring, and that 31% feel fully capable of securing agent AI systems. Cisco also referenced Mythos as an example of new model capabilities changing the threat landscape and cited work across Secure Access, Duo Agentic Identity, firewall inspection, MCP gateways, and integration with Splunk for SOC automation.

Per the blog post, Cisco said it has expanded its Zero Trust Access architecture with Secure Access and Duo Agentic Identity capabilities to discover and authorize agents and their actions. The post describes an "AI Defense" portfolio that includes tools to scan and harden models, gateways to inspect AI traffic, and automated SOC workflows through Splunk. Cisco framed these capabilities as aimed at improving visibility, governance, and response for agentic systems; the post did not publish acquisition financial terms or an integration timeline.

Editorial analysis: The acquisition announcement fits a broader industry pattern where established security vendors are incorporating non-human identity controls and runtime governance into Zero Trust frameworks. For practitioners, the move highlights demand for tooling that can: 1) identify machine identities and agent endpoints, 2) enforce action-level authorization and monitoring, and 3) integrate model-level telemetry into detection and response pipelines.

For practitioners: Watch for Cisco product messaging and documentation that clarify how Astrix capabilities map to existing Zero Trust controls, which APIs or telemetry formats will be supported for agent visibility, and whether third-party validation or integrations (for example with SIEMs or policy engines) appear. Also monitor regulatory and standards activity around non-human identity and agent governance, since those will shape enterprise adoption and vendor roadmaps.