Artificial Intelligence AI Ethics and Responsible AI 2 — Questions and Answers
Question 1: What is 'model fairness' in AI, and why does it matter?
- Ensuring the model runs at the same speed for all users
- Ensuring the model's predictions are equitable across demographic groups and do not perpetuate systemic discrimination (Correct answer)
- Making sure the model is open-source
- Ensuring training and test accuracy are similar
Correct answer: Ensuring the model's predictions are equitable across demographic groups and do not perpetuate systemic discrimination
Model fairness ensures that AI decisions do not systematically disadvantage protected groups (e.g., by race, gender, age), promoting equitable outcomes in consequential applications.
Question 2: What is the 'right to explanation' under GDPR in the context of automated AI decisions?
- The right to request a copy of your personal data
- The right of individuals to receive meaningful information about the logic behind automated decisions that significantly affect them (Correct answer)
- The right to delete your data from AI training sets
- The right to opt out of all AI recommendations
Correct answer: The right of individuals to receive meaningful information about the logic behind automated decisions that significantly affect them
GDPR Article 22 grants individuals the right not to be subject to solely automated decisions with significant effects, and to obtain meaningful explanations of the decision logic.
Question 3: What is 'model auditing' in responsible AI practices?
- Reviewing the training code for syntax errors
- Systematically evaluating an AI model for bias, accuracy, safety, and compliance with ethical standards (Correct answer)
- Tracking model inference costs over time
- Updating model weights after deployment
Correct answer: Systematically evaluating an AI model for bias, accuracy, safety, and compliance with ethical standards
Model auditing involves third-party or internal evaluation of AI systems to identify unfair outcomes, errors, compliance gaps, or potential harms before and after deployment.
Question 4: What does 'transparency' require of AI developers and deployers?
- Making all model weights publicly available
- Being open about how AI systems work, what data they use, their limitations, and potential risks (Correct answer)
- Ensuring the AI is available 24/7 without downtime
- Publishing all training datasets online
Correct answer: Being open about how AI systems work, what data they use, their limitations, and potential risks
Transparency requires that AI developers disclose how systems are built, what they can and cannot do, and where they may fail, enabling informed use and external scrutiny.
Question 5: What is 'data minimization' as it relates to responsible AI?
- Using the smallest possible neural network
- Collecting and retaining only the personal data strictly necessary for the intended AI application (Correct answer)
- Reducing the number of training epochs
- Limiting the size of input images
Correct answer: Collecting and retaining only the personal data strictly necessary for the intended AI application
Data minimization reduces privacy risk by limiting data collection and retention to what is genuinely needed, aligning with GDPR and privacy-by-design principles.
Question 6: What is 'human-in-the-loop' (HITL) in AI systems?
- A human manually labeling every model output after deployment
- A design approach where human judgment is integrated into AI decision-making for oversight and correction (Correct answer)
- A requirement that AI code be written by humans, not other AI
- A safety system that shuts down AI when error rates rise
Correct answer: A design approach where human judgment is integrated into AI decision-making for oversight and correction
HITL systems incorporate human review and approval at critical decision points, ensuring that humans can catch errors, override AI judgments, and maintain accountability.
Artificial Intelligence AI Ethics and Responsible AI 2
What is 'model fairness' in AI, and why does it matter?