OneTrust AI Governance

1. 

   AI estate discovery and inventory

   Discovers and catalogs AI models, systems, agents, datasets, vendors, and projects across shifting technical environments. It maps relationships, identifies operational footprints, and assigns explicit lifecycle accountability.

   Use caseA compliance lead catalogs an organization's existing internal and vendor models to see which assets lack a formal review before deployment.

   From onetrust.com/products/ai-governance · image from OneTrust AI Governance
2. 

   Standardized AI risk assessment

   Evaluates risk levels for homegrown and vendor tools using prebuilt templates aligned to frameworks like ISO 42001, the EU AI Act, and NIST. It automates tiering and triggers revalidation when system components or data change.

   Use caseA risk analyst evaluates a newly proposed machine learning use case against the EU AI Act and automatically assigns a risk tier based on data sensitivity.

   From onetrust.com/products/ai-governance · image from OneTrust AI Governance
3. Runtime AI observation and monitoring

   Monitors production model and agent activity across supported cloud platforms to assess performance, quality, and safety signals. It tracks token usage and inspects prompts and responses for sensitive personal details.

   Use caseA security engineer inspects live interaction logs on Amazon Bedrock to discover sensitive personal data leaking through user prompts.

   From onetrust.com/products/ai-governance
4. AI guardrail management and enforcement

   Enforces technical policies directly where models operate to control actions and screen data. It allows teams to block, allow, redact, or redirect inputs and outputs to prevent policy violations.

   Use caseA system administrator applies prebuilt guardrails to filter out sensitive business information from prompt submissions before a public model processes them.

   From onetrust.com/products/ai-governance
5. 

   Agent action governance

   Restricts AI agents to designated purposes, access permissions, and permitted tools across connected and MCP environments. It records auditable event logs and rechecks rules during multi agent interactions.

   Use caseA platform engineer enforces strict MCP policies on an agent to keep it from executing unauthorized external tool actions.

   From onetrust.com/products/ai-governance · image from OneTrust AI Governance

Written from OneTrust AI Governance's own website, checked Oct 16, 2026. Each feature links to the page it comes from; images are OneTrust AI Governance's own.