SecAI+ Network Isolation for AI Environments

1. In a secure AI environment, what does VPC (Virtual Private Cloud) isolation provide?

Logical network boundaries and traffic control

Physical separation of servers

Automatic malware detection

Real-time model performance monitoring

2. In secure AI deployment, how should communication with external APIs be protected?

Through dedicated secure tunnels with certificate validation

Allowing any outbound connection to external services

Using only HTTP connections

Disabling all external API access

3. What is the purpose of egress filtering in AI environment network isolation?

Restricting outbound traffic to approved destinations only

Blocking all incoming connections

Monitoring user login attempts

Optimizing network bandwidth usage

4. How should network isolation be configured for AI systems handling personally identifiable information (PII)?

Separate network segment with strict egress controls and encryption

Open network access for ease of use

Isolated from all other systems completely

No special isolation needed if data is encrypted

5. What does API gateway isolation accomplish in secure AI environments?

Controls and authenticates all external requests to AI services

Automatically trains AI models

Manages database backups

Schedules resource allocation

6. In containerized AI deployments, what network isolation mechanism is commonly used?

Container network interfaces (CNIs) with pod-to-pod policies

Physical network cables

USB connections

Shared memory segments

7. What is the primary benefit of implementing network monitoring in isolated AI environments?

Detecting anomalous traffic patterns and potential security breaches

Increasing model training speed

Reducing cloud infrastructure costs

Simplifying backup procedures

8. How does network isolation protect against data exfiltration in AI environments?

By preventing unauthorized outbound connections from sensitive systems

By encrypting data at rest only

By removing all network connectivity

By requiring passwords for every data access

9. What is the purpose of network access control lists (ACLs) in AI systems?

To define which traffic is allowed or denied between network segments

To manage user login credentials

To monitor GPU utilization

To schedule model training jobs

10. Which protocol should be used for secure remote access to AI deployment infrastructure?

Telnet

SSH with key-based authentication

HTTP

FTP

11. What is the primary purpose of network segmentation in AI deployment environments?

To isolate AI systems and limit lateral movement of threats

To increase network bandwidth for faster model training

To reduce infrastructure costs

To simplify network management protocols

12. What is the role of network segmentation when isolating sensitive AI training data?

It prevents unauthorized access to data networks from other environment zones

It increases data processing speed

It eliminates the need for encryption

It reduces storage requirements

13. How should communication between AI microservices be protected in a secure deployment?

Using HTTP for simplicity

Implementing mutual TLS authentication and encryption

Disabling all inter-service communication

Using shared credentials for all services

14. What is microsegmentation in the context of secure AI deployment?

Dividing the network into smaller, isolated zones with granular controls

Reducing network packet size for efficiency

Compressing network traffic for faster transmission

Fragmenting AI model weights across servers

15. Which component is essential for enforcing network policies between AI model training and inference environments?

Load balancers

Network policy controllers

DNS servers

NTP synchronization services

16. In zero-trust architecture for AI deployment, what must be continuously verified?

Only user identity at login

Every access request, regardless of network location

Network traffic once per day

Device manufacturer credentials

17. What does a stateful firewall do differently than a stateless firewall in AI environments?

It tracks connection states and only allows established connections

It operates faster without maintaining state information

It requires less computational overhead

It automatically updates firewall rules based on traffic patterns

18. Network isolation for AI systems typically employs which of the following controls?

Firewalls and access control lists

Virtual private networks and microsegmentation

Intrusion detection systems

All of the above

19. In a secure AI environment, what is a demilitarized zone (DMZ) primarily used for?

Storing sensitive training data

Hosting publicly accessible AI services while protecting internal systems

Running GPU-intensive computations

Managing user authentication credentials

20. Which security architecture principle requires that AI systems operate with minimal required network permissions?

Defense in depth

Zero trust

Principle of least privilege

Defense by obscurity

SecAI+ Network Isolation for AI Environments Quiz