Astrix AI Agents: Secure Compliance Risk Management for Enterprises
▼ Summary
– Astrix Security launched the AI Agent Control Plane (ACP) to deploy secure-by-design AI agents with short-lived, precisely scoped credentials and just-in-time access.
– The solution addresses agentic threats, as 80% of companies have experienced unintended AI agent actions like unauthorized access or data leaks.
– ACP provides a Discover–Secure–Deploy framework for enterprises to adopt agentic AI responsibly, covering homegrown, third-party, or shadow deployments.
– It offers core benefits including auditable actions, least-privilege access, and developer productivity through policy-driven, pre-approved permissions.
– Security admins define granular permission profiles, while developers deploy agents with these policies, all centrally monitored and managed in Astrix’s inventory.
Astrix Security has introduced a new platform called the AI Agent Control Plane (ACP), designed to help enterprises safely deploy and manage AI agents. This solution provides short-lived, precisely scoped credentials and just-in-time access based on least privilege principles, effectively reducing compliance risks and eliminating access chaos across organizational systems.
According to Alon Jackson, CEO of Astrix, agent-related threats have moved from theory to reality. He notes that recent studies indicate 80% of companies have already encountered unintended AI agent behaviors, including unauthorized system access and data exposure. Jackson emphasizes that ACP represents a critical advancement in agent security, enabling businesses to adopt AI technology responsibly while maintaining productivity and control.
Existing Astrix clients already use the platform to identify shadow AI agents and enforce security protocols. With the addition of ACP, organizations can now extend these capabilities to the deployment phase. The complete Discover–Secure–Deploy framework supports large-scale AI adoption, whether agents are developed in-house, sourced from third parties, or operating as unmanaged shadow deployments.
Idan Gour, CTO at Astrix, highlights the importance of Non-Human Identities (NHIs) in AI agent security, referencing OWASP guidelines that stress their role in defining agent permissions. Gour explains that securing AI agents begins with controlling access, and securing access requires managing NHIs, a domain where Astrix has led since 2021. The company’s combination of NHI expertise and AI monitoring allows it to detect shadow agents and identify behavioral risks more effectively than other solutions.
ACP offers three primary benefits: streamlined audits with clear ownership and activity tracking, secure access through least-privilege credentials from the start, and enhanced developer productivity via policy-driven, pre-approved permissions that reduce delays.
The operational process involves three stages. Security administrators begin by defining granular, least-privilege permission profiles specific to each AI agent use case. Developers then deploy agents using their preferred tools, selecting from these pre-approved profiles. Once active, every agent becomes visible within Astrix’s centralized inventory, allowing for real-time monitoring, management, and policy adjustments.
With the launch of ACP, Astrix now offers a comprehensive enterprise solution for AI agent security, completing its end-to-end Discover–Secure–Deploy framework.
(Source: HelpNet Security)
