Most Zero Trust programs were designed around humans, not AI agents. Every link in the traditional Zero Trust chain assumes a **human-verified identity**, an **accountable owner**, **least-privilege access**, and **continuous reverification**. AI agents break these assumptions:

• They often run with **standing credentials** that no one wants to touch.
• They **invoke APIs, hold credentials, and make decisions** without being treated as first-class identities.
• Many organizations **cannot produce a complete inventory** of the AI agents already running in their environment.

The result is what many teams are now facing: a mature Zero Trust program that effectively covers humans, but leaves AI agents largely unexamined. In practice, that means **Zero Trust for only half your workforce**. Cross Identity’s approach is to reimagine AI agents as governed identities:

• Each agent is **discovered across 8 planes**, including those no one registered.
• Each agent is given a **verified, owned, revocable identity**, not just a static credential.
• Lifecycle, certification, and entitlement reviews for agents **flow through your existing IGA**, just like human users.

This closes the Zero Trust gap for the “workforce that never logs in.”

Cross Identity focuses on moving AI agents from **unknown to governed** using a few key capabilities: 1. Comprehensive discovery (8 planes)

• The ARIA component **discovers every agent across 8 planes**, including agents that were never formally registered.
• This creates an **agent census** that acts as a Zero Trust gap analysis, showing you what is currently ungoverned before you decide what to govern.

2. Unified identity governance

• Agents are modeled in **one object model** alongside your human workforce.
• Lifecycle management, certification, and entitlement reviews for agents **run natively through your existing IGA**.
• Each agent gets a **Birth Certificate** that records its origin and a **named human owner**, so there is clear accountability.

3. Live, explainable risk scoring

• Every agent receives a **single, live trust score** that all pillars read from.
• This score is **explainable** and updates as conditions change, supporting continuous Zero Trust decisions.

4. Integrated platform, not bolt-on tools

• Agents flow through the same **NimbleNova-class connector** as your human workforce.
• Xenetra Risk runs on the same **WarChief engine** used for Access Management (AM), Identity Governance & Administration (IGA), and CGF.
• The platform is built as **one identity fabric**, not stitched-together acquisitions.

Together, these capabilities help you move from a partial view of AI activity to a governed, scored, and enforced AI agent ecosystem in your own cloud.

Cross Identity is designed to make enforcement **live in the cloud**, not just on tickets or static reports. 1. Verified, revocable identities for agents

• Each agent gets a **verified, owned identity** instead of a long-lived credential.
• Access can be **revoked quickly**, aligning with Zero Trust principles of least privilege and continuous verification.

2. Fast, cloud-level enforcement

• The **Suspend** capability attaches a **deny-all policy** to the IAM principal in seconds.
• This is **verified live in AWS**, changing the cloud configuration directly rather than just creating a ticket.
• Lifecycle, real-time risk, and enforcement are **converged** in the platform, so decisions and actions stay in sync.

3. Independent, corroborating evidence trails

• You get **two independent, corroborating trails**: Xenetra’s own logs and the cloud’s own records.
• This supports **forensics** and makes it easier to demonstrate what happened, when, and why.

4. Compliance support out of the box

• Controls are mapped to frameworks such as the **EU AI Act**, **NIST AI RMF**, and **ISO 42001**.
• Evidence is available **on demand**, helping you show that AI agents are governed and monitored under recognized standards.

By combining live trust scores, direct cloud enforcement, and strong evidence trails, Cross Identity helps you **rethink Zero Trust** so it consistently covers both humans and AI agents.