Identity emerges as governance layer for agentic AI

If anything is clear about the agentic AI revolution, it is that the process for governing agents at scale hinges on identity. This week, Estonia moved to assign digital identities to AI agents in order to ensure accountability, traceability and a human link. As an early adopter of many digital transformations, it heralds a coming shift that is already transforming the digital identity industry.

Across the spectrum, firms are moving to remodel identity security architecture for the agentic era, as enterprise opens operational workflows to autonomous agents that bring both promise and risk with them.

Token hard gates link human oversight to high risk

Token has announced the expansion of its identity security architecture to protect enterprise AI agents. A release says organizations can place Token biometric hard gates around high-consequence agent actions, including sending money, deleting data, changing access rights, releasing confidential information, modifying production systems, and approving sensitive transactions.

AI agents, says the firm, are being connected to systems of record, financial platforms, cloud consoles, help desk tools, development environments, and customer data – critical digital infrastructure that makes up the fabric of the online world. Whether or not a misstep is malicious or accidental hardly matters.

Token’s system requires an authorized human to approve high-consequence actions using a Token biometric device. “AI agents are becoming part of the enterprise operating system,” says Kevin Surace, CEO of Token. “That is incredibly powerful, but it also means agents need real control points. More AI watching AI is useful, but it is still probabilistic. Biometric assured identity is deterministic. When an action matters, the right human must be physically present and must approve it biometrically. That is the control model enterprises need now.”

“Everyone is moving quickly to deploy AI agents,” says Surace. “The leaders are already realizing that agent autonomy needs assured human control. Token is already there.”

Ping extends Runtime Identity into more paths

Ping Identity has announced integrations with Amazon Web Services (AWS), Google Cloud, and Cloudflare that extend Runtime Identity into cloud and edge environments where AI agents operate.

AI agents invoke tools, call APIs, move across cloud workloads, interact with agent gateways and MCP servers, and reach data and services at the edge. A release says Ping is extending its runtime identity capabilities into those execution paths, bringing identity enforcement closer to the point of action to help organizations maintain real-time visibility and control as AI-driven operations scale.

“Organizations want to move faster with AI, but they can’t afford to lose visibility or control as AI agents begin operating autonomously across cloud and edge environments,” says Andre Durand, CEO of Ping Identity. “These integrations help bring continuous authorization and real-time policy enforcement into the environments where AI agents are being built and deployed.”

Okta and Google a ‘natural fit’

A release from Okta says it is expanding its collaboration with Google Cloud to help joint customers strengthen security and resilience in the AI workforce. A new integration between Okta and Gemini Enterprise Agent Platform extends enterprise-level security and identity governance to AI agents.

For agents built on the Gemini Enterprise Agent Platform, Okta for AI Agents will enable continuous importing and registration in a centralized directory, allowing every agent to be linked to a human owner. As external agents interact with Google services, Google Agent Gateway within the Gemini platform will act as the enforcement point, delegating real-time authentication and authorization to Okta for AI agents.

“Organizations shouldn’t have to choose between the AI and productivity tools their teams want and the security their business requires,” says Okta Chief Product Officer Ely Kahn. “Okta and Google are a natural fit because we pair Google’s leading product suite with an identity layer that can work across the entire modern, AI-powered work stack.”

Keeper Security, Wiz unite discovery, remediation in agentic environments

Keeper Security has announced a new integration with Wiz, a cloud and AI security platform that is now part of Google Cloud. A release says joining the Wiz Integration Network “establishes Keeper as a remediation engine for identity security vulnerabilities discovered by Wiz, closing the loop between cloud security detection and active risk resolution.”

Remediation scope covers human users, machine identities, and AI agents and database accounts.

“Finding a vulnerability is the first half of the battle,” says Craig Lurey, CTO of Keeper Security. “By integrating with Wiz, Keeper helps customers rotate compromised credentials, enforce privileged access management and reduce over-permissioned identities, turning Wiz’s detection power into faster, more decisive risk reduction. This is the future of cloud security – detection and remediation working as one, giving security teams a clear path from vulnerability discovery to resolution.”

Securden offers single agentic control plane

Securden has launched its AI Agent Security and Governance platform, which a release says gives organizations visibility and control over AI agents across endpoints, cloud environments, SaaS applications, MCP servers and connected tools.

“AI agents are rapidly becoming the most powerful non-human identities inside the enterprise,” says Bala Venkatramani, CEO of Securden. “The challenge is that they often operate across multiple environments with excessive privileges, limited visibility, and fragmented governance. We built this capability as part of our Unified Identity Security Platform so organizations can discover, govern, and secure AI agents through a single control plane, with the visibility, controls, and accountability needed to deploy AI at scale.”

Article Topics