AI Risk Assessment Framework Basics

1. What does 'fairness' mean in the context of AI risk assessment?

Treating all inputs equally regardless of context

Ensuring equitable outcomes across demographic groups

Using only objective data sources

Removing all human judgment from decisions

2. True or False: Model explainability is a requirement under the EU AI Act for high-risk AI systems.

True

False

3. True or False: AI governance frameworks are only necessary for large enterprises.

True

False

Submit

7. What is 'explainable AI' (XAI) primarily designed to address?

Making models run faster

The inability to understand how models make decisions

Reducing the cost of data collection

Eliminating bias entirely

8. Which of these represents a compliance risk for AI systems?

Using outdated training data

Violating privacy regulations or contractual obligations

Running models on older hardware

Having fewer than five stakeholders

9. What is the role of 'data governance' in AI risk management?

Storing data on external servers

Ensuring data quality, security, and appropriate usage

Maximizing the size of training datasets

Eliminating the need for data validation

10. Which method is commonly used to assess AI model robustness?

Adversarial testing and stress testing

Counting total data records

Measuring server uptime

Reviewing team qualifications

11. What is the primary purpose of an AI risk assessment framework?

To eliminate all AI system failures

To identify, evaluate, and mitigate risks from AI deployments

To guarantee regulatory approval for all AI models

To increase AI model accuracy above 99%

12. Which of these is NOT typically part of compliance requirements for AI systems?

Data privacy and protection measures

Regular auditing and testing

Guaranteed model accuracy levels

Transparency and documentation

13. What is the purpose of AI model validation in a risk framework?

To confirm the model meets performance and safety standards

To prove the model is always correct

To eliminate the need for monitoring

To reduce computational costs

14. Which stakeholder group is responsible for AI governance oversight?

Data scientists alone

Executive leadership, compliance, and risk management

IT operations only

External regulators exclusively

15. What is 'model drift' in AI systems?

Changes in model performance due to data distribution shifts

Moving the AI model to a different server

Deliberate adjustments to model weights

The model losing relevance after one year

16. Which risk category includes unauthorized access to AI systems?

Model drift risk

Security and cybersecurity risk

Data quality risk

Reputational risk

17. In AI risk assessment, what does 'explainability' mean?

Making AI decisions understandable to stakeholders

Providing written documentation only

Simplifying model architecture

Reducing the number of input features

18. Which regulatory framework is most relevant to AI governance in the European Union?

GDPR and EU AI Act

HIPAA and CCPA

SOX and Basel III

FCA and PRA guidelines

19. What does 'algorithmic bias' refer to in AI risk?

Intentional discrimination by developers

Systematic errors favoring certain groups over others

Using biased training data deliberately

The model's preference for specific algorithms

20. Which of the following is a key component of AI governance?

Model versioning and documentation

Stakeholder accountability and oversight

Regular risk assessments and audits

All of the above

SecAI+ AI Risk Assessment Framework Basics Quiz