AI Ethics and Bias Flashcards
6 cards from real Artificial Intelligence practice questions. Tap to flip, then mark Knew It or Still Learning — missed cards come back until you master them.
Read the first 6 AI Ethics and Bias flashcards as text
What is 'differential privacy' designed to protect?
Answer: The privacy of individuals in datasets by adding calibrated noise to outputs
Differential privacy adds carefully calibrated noise to data or outputs so that individual records cannot be inferred from results.
What is an 'adversarial example' in the context of AI safety?
Answer: An input crafted to fool a model into making an incorrect prediction
Adversarial examples are inputs with small, often imperceptible perturbations that cause AI models to make confident but wrong predictions.
What is 'AI alignment' research primarily concerned with?
Answer: Ensuring AI systems behave in accordance with human values and intentions
AI alignment studies how to ensure advanced AI systems pursue goals that are beneficial and consistent with human values.
Which US agency is primarily responsible for regulating AI in consumer financial products?
Answer: CFPB (Consumer Financial Protection Bureau)
The CFPB oversees AI-driven credit scoring and lending decisions to ensure they comply with fair lending laws.
What does 'debiasing' a dataset mean?
Answer: Correcting imbalances or prejudicial patterns in training data to reduce unfair outcomes
Debiasing involves identifying and reducing skewed representations or prejudicial labels in training data that lead to biased models.
What is 'human-in-the-loop' AI?
Answer: A system where human judgment is incorporated at critical decision points
Human-in-the-loop systems keep a human reviewer involved at key stages to validate, correct, or approve AI decisions.