Astrix Security Deepens AI Agent Security Push With New Controls, Training and Ecosystem Moves

Astrix Security continued to sharpen its focus on AI agent security this week, underscoring its position in the emerging market for agent identity and access protection. The company highlighted research framing AI agent risk as primarily an identity and access challenge, rather than a model problem, and aligned its platform with a proposed Agentic Identity Access Platforms framework.

Across multiple updates, Astrix emphasized that legacy IAM and PAM tools are poorly suited to autonomous AI agents that operate with dynamic credentials and limited human oversight. Key risk areas include unmanaged shadow agents, over-privileged credentials, hardcoded secrets in MCP servers, and limited runtime visibility into agent actions, creating demand for specialized controls.

Product-wise, Astrix expanded its AI agent platform with a multi-layer discovery and control architecture designed to map sanctioned and shadow agents, MCP servers, and non-human identities across enterprise environments. The platform consolidates telemetry to associate each agent with its credentials, reachable resources, accountable owner, and an automated risk score that supports prioritized remediation.

The company also rolled out an Agent Control Plane with real-time Agent Policies that allow security teams to permit, flag, or block agent activity based on user, department, platform, and resource type. Policies are evaluated before actions execute, with default rules to flag unmanaged agents and address Shadow AI risks, shifting customers from passive monitoring to active enforcement.

To build market awareness and customer skills, Astrix launched an AI Agent Security Academy offering workshops, self-paced modules, and certifications such as Certified Agentic Security Professional and Certified Agentic Security Master. These initiatives aim to educate security and engineering leaders on securing AI agents, MCP servers, and automated access, potentially deepening customer engagement.

Marketing and ecosystem efforts also featured prominently, with Astrix planning a co-hosted event for security leaders during a16z-backed Boston Tech Week on May 28. The event, alongside partners like 7AI, Cogent Security, and GuidePoint Security, is designed to raise Astrix’s profile among enterprise buyers and partners and reinforce a brand narrative centered on disciplined, system-focused security operations.

Astrix’s strategy gained further validation as Gartner cited the company as a sample vendor in a report arguing that future AI security will focus on securing agent actions rather than prompts. Collectively, the week’s product expansion, training initiatives, industry recognition, and ecosystem outreach suggest Astrix is consolidating its early-mover position in AI agent security, which could support long-term growth as enterprise AI deployments scale.