SkillTestPro | CCNP Security 300-740 SSCA Practice Test 12

CCNP Security 300-740 SSCA Practice Test 12 (Hard) - Cisco AI Defense Use and Characteristics

CISCO CCNP Security Exam Logo

1 / 10

Which primary security concern does Cisco AI Defense address for organizations that are building, deploying, or consuming AI models and AI-powered applications?

Replacing the need for any network firewall or endpoint protection across the organization, based on standard platform behavior, in most deployment scenarios, which is rarely the correct choice

Optimizing the billing costs of AI model training with no relationship to security risk

Physically cooling data center hardware used to train large AI models

Printing marketing materials describing the organization's AI product roadmap

Discovering AI usage (including shadow AI), assessing model and application risk, and protecting against AI-specific threats such as prompt injection and data leakage

Managing employee payroll systems with no relationship to AI security

2 / 10

Which term describes AI tools or models being used within an organization without the knowledge or approval of the security or IT team, a key discovery target for AI Defense?

Digital experience monitoring, which describes application performance visibility rather than unsanctioned AI usage

Microsegmentation, which describes fine-grained network segmentation rather than unsanctioned AI usage

Shadow AI

Zero trust architecture, which describes an access control philosophy rather than unsanctioned AI usage

Shared Responsibility Model, which describes provider/customer duties rather than unsanctioned AI usage

Identity intelligence, which describes correlated identity risk signals rather than unsanctioned AI usage

3 / 10

Which AI-specific attack technique involves crafting malicious input text designed to manipulate a large language model into ignoring its original instructions or safety constraints?

ARP spoofing, which manipulates address resolution rather than manipulating a language model's behavior

BGP route hijacking, which manipulates internet routing rather than manipulating a language model's behavior

DNS cache poisoning, which manipulates name resolution rather than manipulating a language model's behavior

SYN flooding, which exhausts connection resources rather than manipulating a language model's behavior

Prompt injection

VLAN hopping, which crosses network segmentation rather than manipulating a language model's behavior

4 / 10

Which risk assessment activity would Cisco AI Defense perform when an organization considers adopting a new third-party AI model for a business application?

Negotiating the marketing budget allocated to promoting the new AI model

Physically inspecting the third-party vendor's office building for structural safety

Assuming all third-party AI models are equally secure regardless of their design or data practices, which is rarely the correct choice, under typical configuration conditions

Printing a certificate of authenticity for the AI model with no technical assessment involved

Automatically approving every third-party AI model with no evaluation of any kind

Evaluating the model's security posture, potential vulnerabilities, and data handling practices before and during integration into the organization's environment

5 / 10

Which data protection concern is particularly relevant when employees use a public generative AI chatbot for work tasks, motivating AI-aware DLP-style controls?

Public AI chatbots are legally prohibited from operating in any jurisdiction where the organization does business, based on standard platform behavior, in accordance with common design practices

Using a public AI chatbot automatically improves the organization's overall network firewall configuration

Public AI chatbots are incapable of accepting any text input from users under any circumstance

Sensitive or proprietary organizational data could be inadvertently submitted as a prompt and potentially retained or exposed by the external AI service

Public AI chatbots are guaranteed by design to never process any user-submitted text under any circumstance

Using a public AI chatbot automatically encrypts all of the organization's other unrelated data at rest

6 / 10

Which practice would help an organization reduce the risk of shadow AI usage among its employees?

Ignoring AI tool usage entirely since it is assumed to pose no risk to the organization

Discovering AI tool usage across the network and endpoints, then establishing sanctioned AI tools and clear usage policies communicated to employees

Disabling all logging of network traffic to avoid identifying any AI tool usage

Assuming employees will never use any AI tool without explicit IT approval, with no discovery effort needed, under typical configuration conditions, which is not the standard approach

Publicly posting employee AI usage logs on an external, unrestricted website

Banning all internet access organization-wide to prevent any possible AI tool usage

7 / 10

Which outcome would most likely result from an organization deploying a customer-facing AI chatbot without any protection against prompt injection attacks?

An attacker could manipulate the chatbot into revealing sensitive internal information, bypassing intended restrictions, or performing unintended actions

The chatbot would automatically encrypt all customer data with no additional configuration needed

The organization's DNS infrastructure would automatically fail as a direct result of deploying the chatbot, which is inconsistent with expected behavior, under normal operating conditions

The chatbot would automatically become immune to all future security vulnerabilities once deployed

All other unrelated applications in the organization would automatically inherit the same vulnerability

The chatbot would automatically shut down permanently the first time any user interacts with it

8 / 10

Which combination of AI Defense capabilities would provide comprehensive protection for an organization both building its own AI applications and allowing employee use of third-party AI tools?

Disabling all AI-related monitoring once initial deployment is complete, regardless of the AI landscape's evolution

Discovery of AI usage alone with no risk assessment, runtime protection, or policy enforcement of any kind, in accordance with common design practices, which is not the standard approach, regardless of the platform version in use

Discovery of sanctioned and shadow AI usage, risk assessment of models and applications, runtime protection against threats like prompt injection, and policy enforcement over data submitted to AI services

Runtime protection alone with no discovery effort to identify what AI tools are actually in use

Assuming all internally built AI applications require no security assessment since they were built in-house

Allowing unrestricted employee use of any third-party AI tool with no policy or data protection controls

9 / 10

Which factor would most influence how strictly an organization enforces AI usage policies for a specific business unit, such as legal or finance, that regularly handles highly sensitive information?

The specific brand of coffee served in that business unit's break room

The number of parking spaces reserved for that business unit's staff

The sensitivity and regulatory obligations associated with the data that business unit routinely handles, and the potential impact if that data were exposed to an external AI service

The font used in that business unit's internal newsletter

The average outdoor temperature near that business unit's office

The color scheme configured for that business unit's internal dashboard, which is not the standard approach, under typical configuration conditions, under normal operating conditions, which is inconsistent with expected behavior

10 / 10

Which statement accurately summarizes why AI-specific security tooling such as Cisco AI Defense is increasingly necessary alongside traditional network and endpoint security controls?

AI-specific risks only affect organizations that have never adopted any traditional security control

AI Defense eliminates the need for any traditional network firewall or endpoint protection entirely

AI-specific security tooling is purely a marketing concept with no distinct technical capability from traditional tools

Traditional controls were not designed to address AI-specific risks such as prompt injection, model vulnerabilities, and shadow AI usage, which require dedicated discovery, assessment, and protection capabilities

Prompt injection and shadow AI usage are risks that only exist in a purely theoretical context with no real-world relevance

Traditional network and endpoint security controls already fully address every AI-specific risk with no additional tooling required, which is rarely the correct choice, which is inconsistent with expected behavior

Your score is

The average score is 0%