Checklist

AI Agent Guardrail Checklist by Risk Tier

An AI agent guardrail checklist verifies that identity controls, permission scoping, tool-call restrictions, monitoring, and audit logging are matched to an agent's risk tier, rather than applying the same controls uniformly regardless of autonomy, data sensitivity, or tool access scope.

Risk Tiering Comes Before Guardrail Selection

NIST's AI Risk Management Framework organizes risk management activity around four functions: Govern, Map, Measure, and Manage. The Map function requires identifying context of use before controls are chosen, and NIST's Generative AI Profile extends this by flagging autonomous operation and expanded system access as risk factors warranting additional management action. In practice, this means an AI agent's risk tier should be established using a small set of criteria before any guardrail is selected: the agent's autonomy level, the sensitivity of data it can access, the scope of tools or APIs it can call, and its reach into external systems. Enterprises that skip this step tend to apply a single guardrail set across all agents, which produces two failure modes at once: low-risk agents carry unnecessary control overhead, and high-risk agents operate with guardrails designed for a much narrower blast radius than they actually have.

What Actually Changes Between Tiers

The technical controls that differentiate risk tiers cluster into five categories: identity verification, permission scoping, tool-call governance, monitoring depth, and audit logging. OWASP's Top 10 for LLM Applications names Excessive Agency as a distinct risk, describing harm that results from granting an agent more functionality, permissions, or autonomy than its task requires, and separately addresses insecure plugin and tool integration design. CISA and NSA joint guidance on deploying AI systems securely recommends least-privilege access and continuous monitoring as baseline controls. NIST SP 800-53's AC-6 (least privilege) and AU control family are commonly adapted as the underlying technical baseline for scoping agent permissions and defining audit logging depth. None of these sources prescribe a fixed tier taxonomy, so the tiering scheme itself is an organizational decision, but the controls that should scale with tier are consistently the same five categories.

Common Coverage Gaps When Tiering Is Skipped

  • Static API keys used for tool connections that should require OAuth-based or scoped authorization, particularly where Model Context Protocol servers expose tools or data
  • Tool access left unrestricted at design time, only discovered as excessive agency after an incident or review
  • High-risk controls such as per-call interception applied inconsistently, leaving some high-autonomy agents under-monitored
  • Monitoring and audit logging built to a single standard rather than scaled to the AU-control depth appropriate for the agent's tier
  • Risk tier reassessed only at initial deployment, without revisiting classification as an agent's tool access or autonomy expands

Three Risk Tiers, Three Guardrail Baselines

Risk tier is determined by autonomy level, data sensitivity, and tool or system reach, not by the agent's function or business owner.

Risk tier Profile
Tier 1: Low-Risk Limited autonomy, internal read-only tool access, low data sensitivity
Tier 2: Medium-Risk Task-scoped autonomy, allow-listed internal tools, moderate data exposure
Tier 3: High-Risk Broad autonomy, external tool or system reach, sensitive data access

Guardrail Checklist by Risk Tier

Use this checklist to verify that identity, permission, tool-call, and audit controls are appropriately scoped for each agent's assigned tier, before and after deployment.

Tier 1: Low-Risk baseline

  • Standard IAM-based identity, no elevated or persistent service credentials
  • Default role-based permissions consistent with existing application access policies
  • Tool access limited to read-only or a single low-sensitivity internal system
  • Standard application-level logging without session-level detail
  • Log retention aligned to existing IT policy, no dedicated audit review cycle

Tier 2: Medium-Risk baseline

  • Service-specific credentials with defined expiration, not shared or static keys
  • Permissions scoped to the specific task, with no standing elevated access
  • Tool access restricted to an allow-listed set of internal tools, no external write access
  • Session-level monitoring covering agent actions, not just request/response logs
  • Action logs retained on a schedule consistent with NIST AU-family baseline expectations

Tier 3: High-Risk baseline

  • OAuth-based or equivalent scoped authorization for tool connections, no static API keys
  • Least-privilege permissions scoped per action, with explicit consent for sensitive operations
  • Tool-call interception and allow-listing enforced per call, not only at session start
  • Real-time session and action-level monitoring with review capability
  • Audit logs aligned to NIST 800-53 AU controls, retained and reviewable for incident response

Verify Guardrail Coverage Before and After Deployment

Trussed AI provides runtime governance for enterprise AI agents, including agent identity, permission scoping, tool approval workflows, and audit logging enforced at runtime rather than defined only in policy documents.

Explore Runtime Governance