Artificial Intelligence AI Ethics and Bias 1 — Questions and Answers
Question 1: What is 'algorithmic bias' in AI systems?
- The tendency of models to overfit training data
- Systematic errors in AI outputs that unfairly favor or disadvantage certain groups (Correct answer)
- High variance in model predictions
- Incorrect hyperparameter settings
Correct answer: Systematic errors in AI outputs that unfairly favor or disadvantage certain groups
Algorithmic bias occurs when AI systems produce systematically unfair outcomes, often reflecting biases present in training data.
Question 2: What does 'fairness' in AI typically require?
- Equal model accuracy across all users regardless of group
- Equitable treatment and outcomes for individuals across demographic groups (Correct answer)
- Models that are free from any statistical patterns
- Identical outputs for all inputs
Correct answer: Equitable treatment and outcomes for individuals across demographic groups
AI fairness aims to ensure that model decisions do not systematically disadvantage individuals based on protected attributes.
Question 3: What is 'explainability' (or interpretability) in AI?
- The ability to run models on edge devices
- The degree to which humans can understand and trace how an AI makes decisions (Correct answer)
- A model's accuracy on held-out test data
- The speed at which a model produces outputs
Correct answer: The degree to which humans can understand and trace how an AI makes decisions
Explainability refers to how transparently an AI system's reasoning can be understood by humans.
Question 4: What is 'data privacy' in the context of AI model training?
- Encrypting model weights
- Ensuring that personal data used in training is handled, stored, and used in accordance with privacy laws (Correct answer)
- Using only public datasets
- Removing duplicate records from training data
Correct answer: Ensuring that personal data used in training is handled, stored, and used in accordance with privacy laws
Data privacy in AI means protecting individuals' personal information used during model training from unauthorized access or misuse.
Question 5: What is the 'right to explanation' in AI regulation?
- A requirement for AI companies to publish source code
- The right of individuals to receive a meaningful explanation of automated decisions affecting them (Correct answer)
- The right to request model retraining
- A requirement for AI models to be open source
Correct answer: The right of individuals to receive a meaningful explanation of automated decisions affecting them
The right to explanation allows people affected by automated decisions to understand the reasoning behind them.
Question 6: What is 'model transparency' in responsible AI?
- Making model weights publicly downloadable
- Openly documenting how a model works, what data it was trained on, and its limitations (Correct answer)
- Ensuring the model has no parameters
- Publishing all training code without exception
Correct answer: Openly documenting how a model works, what data it was trained on, and its limitations
Model transparency involves disclosing key information about the model's design, training data, and performance to enable accountability.
Artificial Intelligence AI Ethics and Bias 1
What is 'algorithmic bias' in AI systems?