AI Security Atlas › All entries › Autonomous Agent & MCP Security

GuardrailFixes high risks

Agent Behavioral Monitoring & Kill Switch

Step-level telemetry, scope-drift detection, and a tested kill switch that halts agents and revokes their credentials.

Impact

Bounds how long a drifting or compromised agent can act.

Why this severity

Guardrails take the highest severity of the risks they mitigate: No Behavioral Monitoring or Kill Switch (High), Rogue or Compromised Agent (High), Goal Alignment Drift (High).

Method: OWASP Risk Rating Methodology

Related

Mitigates

• No Behavioral Monitoring or Kill Switch Vulnerability · High severity
• Rogue or Compromised Agent Threat · High severity
• Goal Alignment Drift Threat · High severity

Remediation

1. Stream agent telemetry to the SIEM with agent and session IDs.
2. Compare each action to the authorized objective and scope.
3. Block agents from granting themselves new permissions.

How to verify

• Scope-drift fixtures trigger alerts in staging.
• Quarterly kill-switch drill results are recorded.

Framework mapping

OWASP Top 10 for Agentic Applications 2026

• ASI10:2026 Rogue Agents
• ASI01:2026 Agent Goal Hijack

MITRE ATLAS

• AML.M0024 AI Telemetry Logging
• AML.M0038 AI Agent Scope Drift Detection
• AML.M0037 AI Agent Authority Expansion Controls

NIST AI RMF 1.0

• MANAGE 2.4 Mechanisms exist to supersede, disengage, or deactivate AI systems
• MANAGE 4.1 Post-deployment monitoring, incident response, and change management