Phase 125: Agent Safety and Security
Phase 125 of the AI Encyclopedia — Agent Safety and Security. Topics 2481–2500.
This phase covers Agent Safety and Security. Below are the 20 concepts grouped under this phase — each is a future article in the Insightful AI World encyclopedia.
2481 Agent Safety
2482 Agent Alignment
2483 Agent Guardrails
2484 Agent Permission Boundaries
2485 Tool Abuse Prevention
2486 Prompt Injection against Agents
2487 Indirect Prompt Injection
2488 Data Exfiltration Risk
2489 Autonomous Action Risk
2490 Sandboxing Agents
2491 Human Approval Gates
2492 Rate Limits
2493 Rollback Mechanisms
2494 Agent Monitoring
2495 Agent Red Teaming
2496 Agent Incident Response
2497 Agent Policy Enforcement
2498 Safe Tool Use
2499 Agent Trust Calibration
2500 High-risk Agent Governance