AI Security & Governance Technical Specifications

  1. AI Security & Governance
  2. /
  3. Technical Specifications

Technical requirements for deploying AI Agent Governance & Security: how the product is hosted, what it needs from your cloud platforms, and what the endpoint agent supports.

Architecture and Deployment

Area Detail
Deployment model SaaS
Hosting AWS
Supported AI agent platforms Amazon Bedrock, Amazon Bedrock AgentCore, Microsoft Copilot Studio, and Microsoft Foundry
Agent requirement Agentless for cloud platform discovery. The Securden endpoint agent is required to detect AI activity on managed devices and route it to the gateway.
Cloud platform permissions Permissions to integrate the gateway and to read AI agent logs. Specific permissions vary by platform.
Gateway Hosted by Securden. No customer-side installation required.
Required outbound connectivity from endpoints HTTPS 443 outbound. No inbound firewall rules required.

Endpoint Agent

Area Detail
Function Discovers AI applications, browser-based AI, AI coding assistants, and other AI activity on managed devices, and routes it to the gateway for policy enforcement
Supported operating systems Windows 10 and later; Windows Server 2019 and later
Deployment methods Group Policy, or manual installation
Proxy support Supported

Security, Access, and Integrations

Area Detail
Supported browsers Chrome, Edge, Firefox, and Safari
Platform authentication Local accounts, Active Directory, Microsoft Entra ID, LDAP, and RADIUS
Single sign-on SAML 2.0 for Okta, Microsoft Entra ID, ADFS, OneLogin, PingIdentity, Google Workspace, and other SAML identity providers
Multi-factor authentication TOTP apps including Google Authenticator and Microsoft Authenticator, Securden Authenticator, Duo Security, RSA SecurID, YubiKey, email OTP, and RADIUS-based MFA
Encryption at rest AES-256
Encryption in transit TLS 1.2 or later
SIEM integration Supported