Microsoft Azure AI Fundamentals Responsible AI 2 — Questions and Answers
Question 1: What is AI bias and why is it a concern in the context of Responsible AI?
- AI bias is intentional programming to favor certain outputs; it is rarely a concern
- AI bias occurs when models produce skewed or unfair results due to unrepresentative training data or flawed design (Correct answer)
- AI bias refers to the model's preference for faster computations
- AI bias is when users intentionally misuse AI systems
Correct answer: AI bias occurs when models produce skewed or unfair results due to unrepresentative training data or flawed design
AI bias arises when training data or model design reflects societal inequities, causing models to produce unfair or discriminatory outcomes for certain groups.
Question 2: Which Azure tool helps teams detect and mitigate fairness issues in machine learning models?
- Azure Monitor
- Fairlearn (integrated with Azure Machine Learning) (Correct answer)
- Azure Policy
- Azure Blueprints
Correct answer: Fairlearn (integrated with Azure Machine Learning)
Fairlearn is an open-source toolkit integrated with Azure Machine Learning that helps assess and mitigate fairness-related harms in machine learning models.
Question 3: In the context of Responsible AI, what is a 'human in the loop'?
- A developer who writes the AI training code
- A person involved in reviewing, overriding, or approving AI decisions to ensure oversight (Correct answer)
- An end user who interacts with the AI product
- An algorithm that simulates human reasoning
Correct answer: A person involved in reviewing, overriding, or approving AI decisions to ensure oversight
Human in the loop refers to a design approach where humans review and can override AI decisions, ensuring meaningful oversight especially in high-stakes scenarios.
Question 4: What does 'model explainability' mean in Responsible AI practice?
- The ability to predict a model's future accuracy
- The capacity to understand and communicate why a model made a specific prediction (Correct answer)
- The process of optimizing model hyperparameters
- The documentation of a model's architecture
Correct answer: The capacity to understand and communicate why a model made a specific prediction
Model explainability refers to techniques that help stakeholders understand why an AI model produced a particular output or decision.
Question 5: Which practice helps ensure AI systems respect user data under the Privacy & Security principle?
- Training models on publicly available data without anonymization
- Implementing data minimization, encryption, and access controls on AI training and inference pipelines (Correct answer)
- Sharing model weights openly to promote transparency
- Storing all user queries indefinitely to improve the model
Correct answer: Implementing data minimization, encryption, and access controls on AI training and inference pipelines
Privacy & Security best practices include collecting only necessary data, anonymizing training datasets, encrypting data in transit and at rest, and restricting access to sensitive information.
Question 6: What is the purpose of a 'model card' in Responsible AI documentation?
- A business card format for AI researchers to share contact information
- A structured document that describes a model's intended use, performance metrics, limitations, and ethical considerations (Correct answer)
- A credit card used to pay for Azure AI services
- A card game used to teach AI concepts to students
Correct answer: A structured document that describes a model's intended use, performance metrics, limitations, and ethical considerations
A model card is a transparency artifact that summarizes what a model does, how it performs across different groups, its limitations, and the contexts in which it should or should not be used.
Microsoft Azure AI Fundamentals Responsible AI 2
What is AI bias and why is it a concern in the context of Responsible AI?