# OneTrust AI Governance

> Enterprise platform that discovers AI systems, assesses risk, and enforces policy guardrails at runtime across models and agents.

OneTrust AI Governance is a module of the OneTrust AI-Ready Governance Platform that helps security, privacy, and governance teams inventory AI, assess its risk, and enforce policy where AI runs. It covers both homegrown and third-party AI, including models, agents, datasets, and vendors. OneTrust says it aims to turn AI policy into controls that detect shadow AI and prevent sensitive data leaks.

## What It Is

AI Governance software that links an AI inventory, risk assessment, policy management, runtime observability, and guardrail enforcement in one program center. The product page describes a continuous audit trail that follows each AI system from intake through production monitoring.

## How the Workflow Operates

Teams bring AI systems, agents, models, datasets, vendors, and projects into one inventory and assign ownership and lifecycle status. Risk is assessed with EU AI Act, NIST AI Risk Management Framework, and ISO 42001 templates, with automated risk tiering by use case, system, deployment context, or data sensitivity. Configurable intake, approval, attestation, and signoff workflows route reviews to the right owners and produce audit-ready evidence.

## Runtime Observation and Enforcement

The page lists runtime monitoring of evaluation, usage, quality, safety, and guardrail signals across Amazon Bedrock and AgentCore, Microsoft Azure Foundry, Databricks, and Gemini Enterprise Agent Platform, with platform coverage varying by integration. AI interaction logs from Amazon Bedrock and Microsoft Foundry are analyzed to detect PII in prompts and responses. AI Policy Manager defines acceptable use and required controls, and Guardrail Enforcement can block, allow, redact, escalate, or route actions. The AI Guard SDK is a Python SDK that identifies sensitive data in prompts and responses before the model sees it.

## Agents and MCP Governance

Agents can be registered with a defined purpose, permissions, and allowed actions. MCP policies are enforced with audit logs, and multi-agent handoffs are governed. Governance is revalidated as agent behavior, tools, or connected data change.

## Recognition

OneTrust states it was named a Visionary in the 2026 Gartner Magic Quadrant for AI Governance Platforms.

## Features
- AI discovery and inventory of systems, models, agents, datasets, and vendors
- Risk assessment templates for EU AI Act, NIST AI RMF, and ISO 42001
- Automated risk tiering
- Intake, approval, attestation, and signoff workflows
- Runtime monitoring of models and agents
- PII detection in AI interaction logs
- AI Policy Manager
- Guardrail enforcement (block, redact, escalate, route)
- AI Guard SDK for in-line sensitive data protection
- Agent and MCP governance with audit logs
- Audit-ready evidence generation

## Integrations
Amazon Bedrock, Microsoft Azure AI Foundry, Google Vertex, Databricks Unity Catalog, Gemini Enterprise Agent Platform, Jira, Palo Alto Networks

## Platforms
WEB, API, DEVELOPER_SDK

## Pricing
Paid

## Links
- Website: https://www.onetrust.com/products/ai-governance/
- EveryDev.ai: https://www.everydev.ai/tools/onetrust-ai-governance