AI Agent Governance / The library
Find the decision you need to make.
practical guides
with examples and source documents
Foundations
Set the direction
Ownership, frameworks and examples of decisions a team needs to make.
What is AI agent governance?
Understand who should own an AI agent, what it may do, and how to review its actions.
AI governance vs AI security
Understand how governance decisions and security controls work together in an agent deployment.
AI governance examples for agent work
Compare completed decisions for diagnosis, a bounded service change and a declined data export.
Access
Define the authority
Identities, permissions and the systems an agent may reach.
AI agent access control at the action boundary
Decide which agent may perform which action on which resource, and enforce the decision.
Least privilege for AI agents
Limit an agent's authority to the actions, resources and time its task actually needs.
AI agent identity and delegated access
Keep the agent, the person requesting work and the executing service identifiable throughout a task.
Production access for AI agents
Give an agent useful operational access with explicit scope, approval rules and recoverable execution.
Non-human identities and AI agents
Connect service accounts, workload identities and delegated users to the agent work they authorize.
Operations
Keep it accountable
Inventories, monitoring, audit trails and the full agent lifecycle.
Human in the loop AI that people can actually review
Design approval steps that give a person enough context and authority to make a useful decision.
What belongs in an AI agent audit trail
Record agent requests, access decisions and outcomes without copying secrets into your logs.
Incident response for an AI agent
Contain unexpected agent activity, preserve evidence and restore access only after checking the cause.
AI agent monitoring and observability
Connect task progress, permission decisions and executor outcomes to signals an operator can act on.
Build an AI agent inventory
Connect each deployment to its owner, identities, tools, target systems and retirement state.
AI agent lifecycle management
Define the authority and evidence changes when an agent is proposed, released, updated, suspended or retired.
Reviews & policy
Review the evidence
Policies, risk assessments and checks before the next release.
An AI agent governance framework for engineering teams
Build a repeatable review process for agent ownership, access, testing and changes.
Write an AI governance policy for agents
Use a practical policy structure that connects agent permissions to owners and enforcement.
How to assess the risk of an AI agent
Review the harm an agent could cause, the controls that limit it, and the evidence still missing.
How to evaluate AI governance tools
Compare inventory, evaluation, policy decisions and action enforcement using a trial your team can verify.
An AI governance checklist for agent access
Collect ownership, scope, test evidence and unresolved decisions before releasing an agent or expanding its permissions.
Assess AI governance maturity
Use observable capability gaps to choose the next improvement, without inventing a certification score.
Apply the NIST AI RMF to agents
Map an agent review to Govern, Map, Measure and Manage while keeping implementation evidence and limitations explicit.
Evaluate an AI agent before granting access
Test useful task completion, permission boundaries, side effects and recovery with evidence suited to each result.
No guide matches yet.
Try a shorter phrase, choose another topic, or return to the complete library.