CCNA 200-301 AI and operations explained

AI, and Network Operations and Management is the smallest domain of the v2.0 CCNA at 10%, and it is new. It covers the role of agentic AI in network operations, choosing a good prompt for a generative AI system, the main network management approaches, SNMP, Ansible, and interpreting syslog messages. Most objectives are descriptive rather than command-heavy, so a few hours of focused study earns real marks.

Agentic AI and generative AI

Cisco’s description of the v2.0 exam says candidates may need to evaluate output and recommendations from agentic AI and digital network assistants. The ideas to know:

• Generative AI produces text, configuration or explanations from a prompt.
• Agentic AI pursues a goal: it plans steps, calls tools (such as running show commands) and acts on the results, rather than answering one prompt.
• A human stays responsible: you verify recommendations before they change a production network.

Writing a prompt

The objective names the components of a good prompt:

The data classification component matters in practice: pasting a full running configuration with secrets into an external AI service is a security failure, whatever the quality of the answer.

Network management approaches

Questions typically describe a workflow, for example one policy change reaching hundreds of devices, and ask which approach it is.

SNMP

SNMP lets a manager poll and receive notifications from agents. The v2.0 objective asks you to describe its function: monitoring through the MIB, get and set operations, and traps or informs sent by devices. Version 3 adds authentication and encryption; versions 1 and 2c use community strings in clear text.

Ansible

Ansible is agentless: the control node connects over SSH or an API, reads an inventory of hosts and runs playbooks written in YAML. It is declarative in style and designed to be idempotent: running a playbook twice leaves the same result.

Syslog

A message such as %LINK-3-UPDOWN carries severity 3 between the facility and the mnemonic. Setting a logging level sends that level and every lower number, so logging trap warnings sends levels 0 to 4.

Example questions

If you sit v1.1 before February 3, 2027

Everything above follows the v2.0 outline. If your booking is for v1.1 (the version Cisco lists as bookable until February 2, 2027), its Automation and Programmability domain (10%) contains two objectives that v2.0 drops or narrows. v2.0 candidates can skip this section.

REST-based APIs (v1.1 objective 6.5)

Objective 6.5 asks you to describe the characteristics of REST-based APIs: authentication types, CRUD, HTTP verbs and data encoding. It is a “describe” objective, so expect recognition questions rather than code. In a controller-based network, this is the northbound interface: a script or application talks to the controller over HTTP(S) instead of logging in to each device.

The core characteristics:

• Client-server and stateless: each request carries everything the server needs, including credentials, because the server keeps no session for the client between calls.
• Resources identified by URIs, for example /api/v1/network-device.
• CRUD maps to HTTP verbs:

• Authentication types: HTTP basic authentication (username and password, Base64-encoded in the Authorization header, so only safe over HTTPS), a token or API key sent with each request after an initial login, and OAuth, where an authorisation server issues the access token.
• Data encoding: JSON is the usual format on Cisco platforms; XML and YAML are other encodings you may see named. The Content-Type and Accept headers say which one is used.
• Response status codes (defined in RFC 9110): 2xx success (200 OK, 201 Created), 4xx client error (400 Bad Request, 401 Unauthorized, 403 Forbidden, 404 Not Found), 5xx server error.

v1.1 also has objective 6.7, recognising JSON components: objects in curly braces, arrays in square brackets, key/value pairs separated by a colon, and keys and string values in double quotes.

Exam-style pitfall: Base64 is encoding, not encryption. An answer that calls basic authentication secure on its own is wrong; anyone who captures the header can decode the password unless HTTPS protects it. A second trap is 401 versus 403: 401 means the request lacked valid credentials, 403 means the server knows who you are but refuses the action.

For a full walkthrough of one request, idempotency and status codes, see REST API basics. NETCONF and RESTCONF, compared in REST, NETCONF and RESTCONF, are not in either CCNA version.

Ansible and Terraform (v1.1 objective 6.6)

Objective 6.6 asks you to recognise the capabilities of configuration management mechanisms such as Ansible and Terraform. Recognising means telling the tools apart from a short description, not writing playbooks or modules. v2.0 keeps Ansible (see the section above) and folds the rest into infrastructure as code, but no longer names Terraform.

What to recognise for each tool:

Both are declarative in style: you describe the end state and the tool works out what to change, which is what makes repeated runs safe. Both are push-based from a control machine, so nothing has to be installed on the network devices. Terraform reaches a platform through a provider, a plug-in that translates its resources into API calls for that platform, such as a cloud service or a network controller.

The link to the rest of the domain is direct: these tools are how automation changes network management (objective 6.1). One versioned definition replaces many manual CLI sessions, which reduces configuration drift and typing errors.

Exam-style pitfall: do not mix up the file formats and the state model. A question that describes “a plan step that shows proposed changes and a state file tracking what was built” is Terraform; “an inventory of hosts and a YAML playbook sent over SSH” is Ansible. An option claiming either tool needs an agent installed on each switch is wrong for both.

The comparison is covered in depth in Ansible vs Terraform, and Terraform’s place among infrastructure-as-code tools in ARM vs Bicep vs Terraform.

Revision tips

Memorise the syslog table with a mnemonic and the five management approaches with one example each. For prompt questions, pick the answer that is specific, protects sensitive data and keeps a human in the loop.

Ready to test everything? Take the free practice test.