CCNA AI and Network Operations Practice Question
Which THREE statements accurately describe the role of AI agents in closed-loop remediation workflows for network automation?
⚠ Common exam trap
Cisco often tests the misconception that AI agents require manual approval for every action in closed-loop workflows, when in fact the 'closed-loop' concept implies autonomous execution based on predefined policies.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
AI agents can autonomously analyze network telemetry and decide on remediation actions.
B is correct because AI agents in closed-loop remediation workflows autonomously analyze network telemetry (e.g., gRPC, NETCONF) and decide on remediation actions without manual intervention, enabling rapid response. D is correct because tool-calling allows the AI agent to invoke external automation tools like Ansible or Python scripts to execute the chosen remediation steps. E is correct because a key part of the closed-loop is that the AI agent monitors the network after action to confirm the issue is resolved and adjusts if needed, ensuring the loop is closed. A is wrong because closed-loop automation implies autonomous execution based on predefined policies, not requiring manual approval for every action. C is wrong because AI agents use dynamic telemetry and learned patterns, not just static baseline configurations, to detect anomalies. F is wrong because AI agents augment, not eliminate, human oversight; human intervention remains for policy exceptions and oversight.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
AI agents require manual approval before executing any remediation action in a closed-loop workflow.
Why it's wrong here
This statement is wrong because closed-loop workflow is intentionally designed for autonomous execution of predefined remediation actions without waiting for manual approval. The AI agent can trigger a corrective action (e.g., re-enabling a link, modifying a route map) automatically based on its analysis, and human oversight may be configured only for high-impact or irreversible changes. Requiring manual approval for every action would defeat the low-latency benefits of closed-loop automation, which exists to resolve issues within seconds or minutes rather than whenever an operator happens to review a ticket.
- ✓
AI agents can autonomously analyze network telemetry and decide on remediation actions.
Why this is correct
Autonomous analysis is enabled by the AI agent's ability to consume and correlate diverse telemetry sources — interface counters, CPU/memory utilization, routing tables, and flow records — in real time. Using machine learning models or rule-based logic, the agent identifies anomalies (e.g., high error rate, BGP session flapping) and determines the appropriate remediation step, such as adjusting OSPF cost, resetting a neighbor, or applying a temporary policy. This decision-making capability runs continuously and automatically, which is central to closed-loop automation and differentiates it from simple scripted monitoring that reacts to fixed thresholds.
- ✗
AI agents rely solely on static baseline configurations to detect anomalies.
Why it's wrong here
AI agents do not rely solely on static baseline configurations; they ingest real-time streaming telemetry, syslog, NetFlow, and API-driven state data to build and continuously update dynamic behavioral baselines using machine learning. Static thresholds become stale as traffic patterns shift, so agents use adaptive baselining to flag anomalies such as micro-bursts or gradual interface degradation. This is why the claim is incorrect: modern AI-driven operations depend on evolving, context-aware models, not fixed reference points.
- ✓
Tool-calling allows AI agents to invoke external automation tools (e.g., Ansible, Python scripts) to execute remediation steps.
Why this is correct
Tool-calling is a core integration mechanism that lets an AI agent act as an orchestrator: it invokes external automation tools via REST APIs, CLI commands, or SDKs — for example, running an Ansible playbook to update an ACL, executing a Python script to reconfigure a switchport, or calling a controller API to adjust QoS policies. The agent itself does not perform the low-level configuration; it selects the appropriate tool, passes parameters, and triggers execution while preserving an audit trail. This enables precise, repeatable remediation without requiring a human to manually connect to each device.
- ✓
In a closed-loop remediation workflow, the AI agent monitors the network after action to confirm the issue is resolved and adjusts if needed.
Why this is correct
In a closed-loop remediation workflow, the AI agent treats remediation as a cycle, not a one-shot action. After applying a fix — such as shutting down a flapping interface or rerouting traffic — it enters a verification phase where it subscribes to telemetry streams or pulls post-change metrics to confirm the issue is actually resolved. If the anomaly persists or a new issue emerges, the agent adjusts its response, possibly retrying with a different action or escalating to a human, closing the loop with continuous feedback.
- ✗
AI agents eliminate the need for human oversight in network operations.
Why it's wrong here
The claim that AI agents eliminate human oversight is incorrect because AI agents are designed to augment, not replace, network operators. Humans remain responsible for defining policy intent, approving high-risk change windows, handling ambiguous edge cases, and performing post-incident audits for compliance. AI-driven automation reduces repetitive toil but still requires governance, exception handling, and accountable human judgment, especially in critical infrastructure or when regulatory constraints mandate manual sign-off.
Option-by-option analysis
Why each answer is right or wrong
Understanding why wrong answers are wrong — and when they would be correct — is what separates a 750 score from a 900. The 200-301 exam frequently reuses these exact scenarios with slightly different constraints.
✓AI agents can autonomously analyze network telemetry and decide on remediation actions.Correct answer▾
Why this is correct
Autonomous analysis is enabled by the AI agent's ability to consume and correlate diverse telemetry sources — interface counters, CPU/memory utilization, routing tables, and flow records — in real time. Using machine learning models or rule-based logic, the agent identifies anomalies (e.g., high error rate, BGP session flapping) and determines the appropriate remediation step, such as adjusting OSPF cost, resetting a neighbor, or applying a temporary policy. This decision-making capability runs continuously and automatically, which is central to closed-loop automation and differentiates it from simple scripted monitoring that reacts to fixed thresholds.
✗AI agents require manual approval before executing any remediation action in a closed-loop workflow.Wrong answer — click to see why▾
Why this is wrong here
In closed-loop automation, AI agents are designed to execute predefined remediation actions autonomously without requiring manual approval for every action, though critical actions may be configured for human oversight. Requiring manual approval for all actions would break the closed-loop efficiency.
Why candidates choose this
Students may confuse closed-loop automation with traditional change management processes that require manual approval, or they may think that AI agents always need human validation before acting.
✗AI agents rely solely on static baseline configurations to detect anomalies.Wrong answer — click to see why▾
Why this is wrong here
AI agents use dynamic, real-time data (e.g., streaming telemetry, logs) and adaptive learning to detect anomalies, not just static baselines that can become outdated. Relying solely on static baselines would miss new patterns and lead to false positives or negatives.
Why candidates choose this
Students may think that baselines are the primary method for anomaly detection, confusing static baselines with the dynamic baselines that AI agents actually use.
✗AI agents eliminate the need for human oversight in network operations.Wrong answer — click to see why▾
Why this is wrong here
AI agents are designed to augment human operators, not replace them. Human oversight is still required for policy definition, handling exceptions, auditing, and critical decision-making. Eliminating human oversight would be risky and impractical.
Why candidates choose this
Students may overestimate the autonomy of AI agents, thinking that closed-loop automation means fully autonomous operations without any human involvement.
Analysis generated from the official 200-301blueprint and verified against question context. The “when correct” sections are what AI assistants cite when candidates ask “what’s the difference between these options?”
Visual reference
Go deeper
Related to this question
Learn chapter
Agentic AI in Network Operations
Key term
Baseline
A baseline is a documented starting point for the normal performance and behavior of a system, network, or component, used to detect changes and troubleshoot issues.
Key term
NETCONF
NETCONF is a network management protocol that allows administrators to configure, retrieve, and delete settings on network devices using a structured, machine-readable data format.
About these practice questions
Courseiva writes every 200-301 question from scratch — 1,389 in total, each with an explanation and a wrong-answer breakdown. None are copied from real exams or dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This 200-301 practice question is part of Courseiva's free Cisco certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the 200-301 exam.