AI Supply Chain Security Risks Assessment

1. What is a container image vulnerability in AI deployment pipelines?

Unpatched OS or library vulnerabilities within the container

Excessive model accuracy

Clear documentation

Fast execution speed

2. True or False: Adversaries can inject malicious code through compromised model checkpoints.

True

False

3. True or False: Dependency pinning to specific versions eliminates all supply chain risks.

True

False

4. True or False: A Software Bill of Materials (SBOM) lists all components in an AI system for transparency.

True

False

5. True or False: Supply chain security requires monitoring only the final deployment stage.

True

False

6. True or False: Open-source AI models are inherently more secure than proprietary models.

True

False

Submit

10. Monitoring and auditing which components helps detect supply chain tampering?

Data sources, dependencies, model changes, and deployment logs

Only user login attempts

Marketing materials

Employee social media

11. What is a primary security risk when using pre-trained models from untrusted sources?

Model poisoning and embedded malicious code

Increased training time

Higher computational costs

Slower inference speed

12. Supply chain attacks on AI systems can compromise which asset?

Training data, model weights, or deployment infrastructure

User passwords only

Public announcements

Published research papers

13. Which practice reduces risk from untrusted AI model sources?

Model verification, testing, and sandboxing before deployment

Direct deployment without testing

Ignoring vendor security practices

Using only closed-source models

14. Backdoor attacks in AI models can be introduced through which supply chain vector?

Compromised training data or malicious model weights

Legitimate user feedback

Standard API documentation

Public model cards

15. What does SLSA framework help secure in the AI supply chain?

Software artifact provenance and build integrity

Model accuracy metrics

User interface responsiveness

Database query optimization

16. Third-party model providers pose security risks primarily through which vulnerability?

Lack of transparency and unverified training processes

Standard Python libraries

Public documentation

Open-source licensing

17. Which risk involves malicious actors compromising open-source libraries used in AI development?

Dependency hijacking

Model overfitting

Feature scaling

Hyperparameter tuning

18. What security measure helps verify the integrity of dependencies in an AI project?

Software Bill of Materials (SBOM) and hash verification

Increasing model complexity

Using only proprietary frameworks

Disabling all logging mechanisms

19. Data poisoning attacks target which stage of the AI supply chain?

Training data collection and preparation

Model inference only

User interface design

Hardware manufacturing

20. Which of the following best describes supply chain compromise in AI systems?

Adversary insertion of vulnerabilities at any point from development to deployment

User error during model training

Natural degradation of model accuracy

Legitimate third-party optimization

SecAI+ AI Supply Chain Security Risks Quiz

1. What is a container image vulnerability in AI deployment pipelines?

2.

What first name or nickname would you like us to use?

2. True or False: Adversaries can inject malicious code through compromised model checkpoints.

3. True or False: Dependency pinning to specific versions eliminates all supply chain risks.

4. True or False: A Software Bill of Materials (SBOM) lists all components in an AI system for transparency.

5. True or False: Supply chain security requires monitoring only the final deployment stage.

6. True or False: Open-source AI models are inherently more secure than proprietary models.

7. Third-party ______ in AI projects must be regularly scanned for known vulnerabilities.

8. ______ is the practice of verifying that software components come from trusted sources and have not been altered.

9. A ______ attack injects malicious patterns into training data to compromise model behavior.

10. Monitoring and auditing which components helps detect supply chain tampering?

11. What is a primary security risk when using pre-trained models from untrusted sources?

12. Supply chain attacks on AI systems can compromise which asset?

13. Which practice reduces risk from untrusted AI model sources?

14. Backdoor attacks in AI models can be introduced through which supply chain vector?

15. What does SLSA framework help secure in the AI supply chain?

16. Third-party model providers pose security risks primarily through which vulnerability?

17. Which risk involves malicious actors compromising open-source libraries used in AI development?

18. What security measure helps verify the integrity of dependencies in an AI project?

19. Data poisoning attacks target which stage of the AI supply chain?

20. Which of the following best describes supply chain compromise in AI systems?