Beyond Identity has launched an AI agent security platform called Ceros, SiliconANGLE reported on June 16.

Beyond Identity provides an identity and access management (IAM) platform that lets companies manage employees' application access privileges.

The Beyond Identity platform applies multi-factor authentication and blocks login requests from insecure devices. Ceros extends that security control scope beyond human employees to AI agents.

A single large company can have hundreds of AI automation workflows. Even when a suspicious file change occurs, it can be difficult to track which AI agent it came from.

Beyond Identity said Ceros automates the process. It logs all AI agent sessions and also stores data on which user ran a session on which device.

Ceros also monitors external components that AI agents rely on when operating. It can map cloud-based large language models (LLMs), MCP servers and local programs used for automated tasks. Administrators can directly set which agent can use which tool under what conditions.

Jason Casey (재슨 케이시), Beyond Identity's chief executive officer, said, "Today's AI threats require strong identity and device trust at the access point before dangerous agents, users and devices touch the network," adding, "With Ceros, chief information security officers gain full visibility and control across all agents, identities, devices and actions."