Token Security, a recognized leader in Non-Human Identity (NHI) security, has announced two groundbreaking innovations designed to help enterprises effectively discover, manage, and secure the rapidly growing ecosystem of AI agents and machine identities. The company has introduced its AI Discovery Engine for NHIs and launched the Token AI Agent, an advanced conversational interface that allows security teams to interact with the Token Security platform using real-time, natural language commands.
These innovations come at a crucial time as organizations increasingly embrace Agentic AI, deploying autonomous agents and complex workflows at unprecedented scale. Existing security tools often fall short of providing the visibility and responsiveness required in this evolving landscape. Token Security’s AI-native solutions are specifically designed to bridge this gap, enabling organizations to stay ahead of emerging security challenges.
Pioneering AI Discovery for Agentic AI Environments
Token Security’s AI Discovery Engine is the industry’s first solution tailored to uncover the full scope of AI-related NHIs within enterprise environments. As the use of AI agents expands, gaining complete visibility into the identities and workloads interacting with AI services becomes essential.
With the AI Discovery Engine, organizations benefit from:
-
Comprehensive AI Visibility: Discover all NHIs and workloads utilizing AI services, including AI connectors.
-
Agent-Aware Intelligence: Identify which AI agents are using specific human or non-human identities, credentials, or systems.
-
AI Agent Insights: Understand how AI agents interact across different environments, platforms, and cloud services.
“This isn’t just another discovery and inventory tool. We are providing the foundation for secure AI transformation,” said Itamar Apelblat, Co-Founder and CEO of Token Security. “The agentic future demands dynamic AI discovery and agile security, combined with context from across your ecosystem, to put you in control of your organization’s agentic AI infrastructure and employee AI adoption.”
Also Read: Okta Adds Cross App Access to Secure enterprise AI Agents
Introducing Token AI Agent for Conversational Security Intelligence
Building on its innovative Model Context Protocol (MCP), Token Security has unveiled the Token AI Agent—a natural language, stream-based interface that allows security, Identity and Access Management (IAM), and development teams to interact directly with their security environment.
With Token AI Agent, teams can effortlessly ask real-time questions and receive actionable insights, such as:
-
“Which identities haven’t rotated secrets in 90 days?”
-
“What secrets are exposed in my dev environments?”
-
“Who owns this service account, and what does it access?”
-
“Which services or workloads are consuming my identities?”
-
“What are the ‘top 5’ riskiest NHIs?”
-
“Generate a script to resolve the top 5 riskiest NHIs.”
The AI Agent provides dynamic querying capabilities across all key platform data layers—including Inventory, NHI Security Posture Management, Secrets, and NHI Threat Detection and Response—delivering immediate insights, explanations, and guided remediation options.
Enabling Scalable Governance for Agentic AI
Together, Token Security’s AI Discovery Engine and AI Agent offer a holistic solution for managing and securing AI-native environments. From uncovering hidden AI agent sprawl to enabling natural language queries for security operations, Token empowers organizations to maintain speed and security in an increasingly AI-driven world.
“With these new capabilities, we’re enabling the shift from fragmented visibility to real-time, agent-driven intelligence to rapidly identify and mitigate critical security risks,” said Nissim Pariente, Chief Product Officer at Token Security. “The Token Security platform is empowering Security teams to see what’s happening within agentic AI infrastructure, identify AI Application usage and shadow AI Applications, understand critical NHI dependencies, and start implementing and enforcing organizational AI security policies.”
