What is AI ethics primarily concerned with?
a) Creating more advanced algorithms
b) Ensuring the responsible use of AI technologies
c) Developing faster hardware for AI
d) Increasing the complexity of machine learning models
Which of the following is an example of bias in machine learning models?
a) A model that always produces the same result
b) A model that gives better predictions for one demographic over another
c) A model that is unable to learn from new data
d) A model that improves over time
How can bias be introduced into machine learning algorithms?
a) Through biased training data
b) By designing more complex algorithms
c) By reducing the size of the dataset
d) By using only numerical data
Which of the following is an example of ethical concern in AI?
a) Automation of tasks
b) AI models used for decision-making in criminal justice
c) Using AI for natural language processing
d) Creating autonomous vehicles
What does “algorithmic fairness” in AI refer to?
a) Making AI systems as powerful as possible
b) Ensuring AI decisions are unbiased and do not disproportionately harm any group
c) Ensuring AI models are easily interpretable
d) Developing algorithms that require no data
What is the primary risk of using biased data in machine learning?
a) Faster model training
b) Ethical issues and unfair outcomes for certain groups
c) Increased accuracy across all groups
d) Decreased computational cost
Why is it important to consider diversity when collecting training data for machine learning models?
a) To improve model performance
b) To ensure the model generalizes well across different groups
c) To make the model more computationally efficient
d) To make the model faster
What can be done to minimize bias in AI models?
a) Use more training data from diverse sources
b) Train models for a longer period
c) Reduce the size of the training dataset
d) Use only numerical data
Which of the following is an ethical issue associated with facial recognition technology?
a) It is too expensive to develop
b) It may infringe on privacy rights and lead to mass surveillance
c) It cannot be used for law enforcement purposes
d) It requires too much computational power
What does “explainability” in AI refer to?
a) The ability to explain the hardware behind an AI system
b) The clarity with which the AI’s decision-making process can be understood by humans
c) The ability to make AI decisions in real-time
d) The accuracy of an AI model’s predictions
What is a potential negative impact of AI on job markets?
a) AI may create more jobs than it eliminates
b) AI may displace human workers, especially in repetitive and manual tasks
c) AI always improves job satisfaction
d) AI never requires human oversight
What does “data privacy” in AI ethics concern?
a) Using large datasets without considering privacy laws
b) Ensuring personal and sensitive data is protected during AI model training and use
c) Training AI models on as much data as possible
d) Using data for commercial purposes without restrictions
Which of the following is a method for detecting bias in AI models?
a) Using complex neural networks
b) Ensuring models are trained on equal amounts of data for all groups
c) Testing the model on a single demographic group
d) Ignoring feedback from diverse user groups
How can AI ethics be incorporated into the design of AI systems?
a) By using only automated decision-making systems
b) By involving diverse teams in the design process and setting clear ethical guidelines
c) By reducing the transparency of AI decision-making
d) By eliminating human oversight in AI decisions
Which of the following actions can be taken to mitigate discrimination in AI systems?
a) Use only numerical data
b) Regularly audit and review AI models for fairness
c) Train models on a limited dataset
d) Eliminate human input from the decision-making process
Why is transparency important in AI models?
a) It helps users understand and trust the decisions made by AI systems
b) It reduces the need for model optimization
c) It speeds up the training process
d) It eliminates bias from the data
What is one of the primary challenges in ensuring fairness in AI models?
a) It is difficult to ensure every model is 100% accurate
b) It is hard to eliminate human oversight
c) It is difficult to obtain diverse and representative training data
d) It is easy to create algorithms that are bias-free
Which of the following is a common ethical issue with AI decision-making systems in healthcare?
a) AI systems never need human input
b) AI systems can make biased decisions, affecting healthcare outcomes for certain groups
c) AI systems always improve accuracy over time
d) AI systems only work in high-resource settings
What role do regulatory frameworks play in AI ethics?
a) They control the algorithms used in AI models
b) They provide guidelines to ensure AI development is safe, fair, and respects privacy
c) They determine the cost of AI systems
d) They help reduce AI’s computational power
Which of the following is an important principle in AI ethics?
a) Autonomous decision-making without human supervision
b) Fairness, accountability, and transparency in AI systems
c) Using as little data as possible in model training
d) Removing human biases from algorithms by eliminating all data
What is a potential consequence of biased AI in hiring practices?
a) More diverse hires across all sectors
b) A fairer, more efficient hiring process
c) Discriminatory hiring practices that disadvantage certain demographic groups
d) Increased employee satisfaction
What is “unintended bias” in AI models?
a) A bias that is deliberately built into a model
b) A bias introduced when the data used for training is not diverse or representative
c) A bias that disappears as the model improves
d) A bias that makes the model more efficient
What is the impact of AI on privacy concerns?
a) AI ensures complete privacy for users
b) AI systems can potentially infringe on privacy by misusing personal data
c) AI eliminates all privacy concerns in data sharing
d) AI has no impact on privacy issues
How can AI systems be made more ethical?
a) By using only raw, unstructured data
b) By eliminating all forms of human supervision
c) By incorporating fairness, transparency, and accountability during the design process
d) By optimizing models to perform better on a narrow dataset
What is a key challenge in applying AI ethics in real-world applications?
a) AI systems always operate without errors
b) Ethical standards are universally accepted and easily implemented
c) Balancing the technological capabilities of AI with ethical considerations and human rights
d) There is no need for ethical oversight in AI development
What is a potential solution to address biases in AI training data?
a) Use random data without considering the source
b) Regularly update the data to reflect diverse groups and contexts
c) Limit the data collection to a single region
d) Remove all subjective human feedback from the training process
Which of the following is a common practice in mitigating AI biases?
a) Training AI on large, diverse, and representative datasets
b) Using AI only for automated tasks with no human oversight
c) Ignoring user feedback
d) Using only historical data from a single demographic group
