HomeSubjectsUniversityBlogAbout

AI Ethics, Security & Privacy

Topic in AI / Machine Learning & Data Analytics

210 total MCQsShowing 30 with explanations10 Easy10 Medium10 Hard
Start Timed Quiz on AI Ethics, Security & PrivacyBack to AI / Machine Learning & Data Analytics

About This Topic

AI ethics studies how to build AI that is fair, transparent, accountable and safe, while AI security and privacy protect those systems and their data. Questions define algorithmic bias and its sources in skewed training data, and fairness criteria such as demographic parity and equalized odds, including the distinction between equality and equity. Explainability items mention LIME, SHAP, the right to explanation and algorithmic recourse. Security questions cover adversarial examples, data poisoning and model inversion attacks, while privacy items address differential privacy, federated learning and GDPR principles. AI safety topics such as alignment and the capability-control trade-off also appear.

Below are 30 practice questions from a pool of 210 AI Ethics, Security & Privacy MCQs, one of 17 topics in AI / Machine Learning & Data Analytics. Each shows the correct answer with an explanation; when you are ready, take a timed quiz to test recall under exam conditions.

Practice Questions

Each question below shows the correct answer with a full explanation. Use these to build conceptual understanding before attempting a timed quiz.

AI Ethics, Security & PrivacyEasy

Q1. What is an adversarial attack on AI?

  1. A.Deliberately crafted inputs designed to fool AI systems✓ Correct
  2. B.A physical attack on the server hardware equipment
  3. C.A network attack on the communication infrastructure
  4. D.A SQL injection attack on the database system

Explanation

Adversarial attacks use carefully crafted inputs (e.g., slightly modified images) that are designed to cause AI models to make incorrect predictions.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q2. What is algorithmic fairness?

  1. A.Making algorithms process data faster through hardware optimization
  2. B.Reducing the overall storage size of trained model files
  3. C.Using more diverse and larger training datasets for models
  4. D.Ensuring algorithms make decisions without discrimination on protected attributes✓ Correct

Explanation

Algorithmic fairness ensures that AI algorithms do not produce biased outcomes that unfairly discriminate against individuals based on protected attributes.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q3. Why is fairness important in AI?

  1. A.To use less training data during the model building phase
  2. B.To ensure AI systems treat all groups equitably without discrimination✓ Correct
  3. C.To make AI systems process data faster and more efficiently
  4. D.To reduce the financial costs of developing AI systems

Explanation

Fairness ensures AI systems don't systematically disadvantage certain groups based on protected attributes like race, gender, or age.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q4. What is AI bias?

  1. A.Systematic unfairness in AI predictions due to biased data or algorithms✓ Correct
  2. B.Slow processing speed during model inference time
  3. C.Random errors that occur during the model prediction process
  4. D.Hardware failures that cause incorrect output computations

Explanation

AI bias occurs when AI systems produce systematically unfair results that favor or discriminate against certain groups due to biased training data or algorithmic design.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q5. What is accountability in AI?

  1. A.Automating every aspect without any human involvement
  2. B.Ensuring responsible parties can be identified when AI causes harm✓ Correct
  3. C.Making AI systems completely autonomous without oversight
  4. D.Removing all human oversight from AI decision processes

Explanation

AI accountability means that individuals and organizations can be held responsible for the decisions and impacts of AI systems they develop or deploy.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q6. What is data privacy?

  1. A.Encrypting only the physical hardware server equipment
  2. B.Making all collected data publicly available and open
  3. C.Deleting all collected data from every storage system
  4. D.Protecting personal information from unauthorized access and use✓ Correct

Explanation

Data privacy involves safeguarding personal and sensitive information, ensuring it is collected, stored, and used in compliance with regulations and individuals' rights.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q7. What is informed consent in AI?

  1. A.Sharing collected data freely with any third party entity
  2. B.Ignoring user preferences about their data usage rights
  3. C.Ensuring people understand and agree to how their data will be used✓ Correct
  4. D.Collecting personal data without any prior user permission

Explanation

Informed consent means individuals are clearly told how their data will be collected and used by AI systems, and they agree to it voluntarily.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q8. What is the purpose of AI regulations?

  1. A.To ensure safe fair and responsible development and use of AI✓ Correct
  2. B.To make AI systems significantly more expensive to build
  3. C.To deliberately slow down all AI research and development
  4. D.To limit all AI capabilities to only narrow simple tasks

Explanation

AI regulations aim to establish guidelines and requirements for the responsible development, deployment, and use of AI systems, protecting individuals and society.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q9. What is transparency in AI?

  1. A.Making AI decision-making processes understandable and open✓ Correct
  2. B.Making AI systems run faster and more efficiently
  3. C.Hiding how AI systems work from all stakeholders
  4. D.Using significantly more data for model training

Explanation

AI transparency means making the decision-making processes, data usage, and limitations of AI systems understandable to stakeholders.

Report an error in this question

AI Ethics, Security & PrivacyEasy

Q10. What is the GDPR?

  1. A.A European regulation for data protection and privacy✓ Correct
  2. B.A deep neural network architecture for images
  3. C.A supervised machine learning training algorithm
  4. D.A general-purpose compiled programming language

Explanation

The General Data Protection Regulation is an EU law that governs data protection and privacy, giving individuals control over their personal data.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q11. What is the difference between individual and group fairness?

  1. A.Group fairness ignores the outcomes of demographic groups
  2. B.Individual fairness treats similar people similarly; group fairness ensures equal outcomes✓ Correct
  3. C.They are completely identical fairness concepts with no differences
  4. D.Individual fairness ignores the treatment of specific individuals

Explanation

Individual fairness requires similar predictions for similar individuals. Group fairness requires statistical parity of outcomes across different demographic groups.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q12. What is federated learning?

  1. A.Sharing all raw training data openly across all participants
  2. B.Training ML models across decentralized devices without sharing raw data✓ Correct
  3. C.Training models exclusively on a single centralized server machine
  4. D.Using only publicly available open-source datasets for training

Explanation

Federated learning trains models across multiple devices or institutions while keeping data local, only sharing model updates, preserving data privacy.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q13. What is differential privacy?

  1. A.A firewall system for blocking unauthorized network access
  2. B.A process of permanently deleting all personal data records
  3. C.A mathematical framework providing privacy guarantees when analyzing data✓ Correct
  4. D.A standard encryption method for securing data files at rest

Explanation

Differential privacy adds calibrated noise to data or queries to ensure that individual records cannot be identified from the analysis results, with provable privacy guarantees.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q14. What is model interpretability?

  1. A.The total file size in megabytes of the serialized model
  2. B.The inference speed at which the model generates predictions
  3. C.The overall accuracy metric of the trained classification model
  4. D.The degree to which humans can understand a model's predictions✓ Correct

Explanation

Model interpretability refers to how well humans can understand and explain why a model makes specific predictions, crucial for trust and regulatory compliance.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q15. What are deepfakes?

  1. A.Corrupted files that cannot be opened or processed
  2. B.AI-generated synthetic media that convincingly replaces a person's likeness✓ Correct
  3. C.Encrypted videos that require a decryption key to play
  4. D.Low-quality images with significant compression artifacts

Explanation

Deepfakes use deep learning to create convincing synthetic media (images, videos, audio) that can manipulate or fabricate someone's appearance or speech.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q16. What is the black box problem in AI?

  1. A.A hardware issue where the server casing is colored black
  2. B.A storage problem where disk capacity runs out unexpectedly
  3. C.Complex models making decisions that cannot be easily understood or explained✓ Correct
  4. D.A networking error where connections time out intermittently

Explanation

The black box problem refers to complex AI models (especially deep learning) whose internal decision-making processes are difficult for humans to understand or interpret.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q17. What is explainable AI (XAI)?

  1. A.AI systems that require significantly less training data
  2. B.AI systems that use significantly larger model architectures
  3. C.AI systems that process data at significantly faster speeds
  4. D.AI systems that provide interpretable explanations for their decisions✓ Correct

Explanation

Explainable AI (XAI) aims to make AI decision-making processes understandable to humans, enabling trust, debugging, and compliance with regulations.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q18. What is the right to explanation under GDPR?

  1. A.Individuals have the right to receive explanations for automated decisions✓ Correct
  2. B.AI companies can keep their algorithms and models completely secret
  3. C.No explanation is ever needed for any automated AI decisions
  4. D.Only government agencies can request explanations of AI systems

Explanation

Under GDPR, individuals affected by automated decision-making have the right to receive meaningful information about the logic involved and potential consequences.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q19. What is data anonymization?

  1. A.Making all personal data publicly available and fully open
  2. B.Encrypting data with standard symmetric encryption only
  3. C.Deleting all data from every storage system permanently
  4. D.Removing or modifying personal identifiers so individuals cannot be identified✓ Correct

Explanation

Data anonymization removes or modifies personally identifiable information so that individuals cannot be re-identified from the dataset.

Report an error in this question

AI Ethics, Security & PrivacyMedium

Q20. What is the trolley problem in AI ethics?

  1. A.A moral dilemma about how autonomous systems should handle potential harm✓ Correct
  2. B.A transportation route optimization problem for logistics planning
  3. C.A data storage capacity issue for managing large files
  4. D.A network routing problem for directing data packets

Explanation

The trolley problem illustrates ethical dilemmas in AI decision-making, such as how autonomous vehicles should prioritize between different potential harms.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q21. What is the alignment problem in AI?

  1. A.Ensuring advanced AI goals and behaviors align with human values and intentions✓ Correct
  2. B.Reducing the computational costs of training large-scale AI models
  3. C.Making AI systems process data at significantly faster inference speeds
  4. D.Training machine learning models to achieve the highest accuracy metrics

Explanation

The alignment problem concerns ensuring that AI systems, especially as they become more capable, pursue goals that are aligned with human values and do not cause unintended harm.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q22. What is membership inference attack?

  1. A.Deleting training data from the model's storage permanently
  2. B.Stealing the complete set of model weights from the server
  3. C.Determining whether a specific data point was used in training a model✓ Correct
  4. D.Changing the model's predictions by modifying its architecture

Explanation

A membership inference attack determines whether a particular data record was part of the model's training dataset, posing privacy risks for sensitive data.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q23. What is model inversion attack?

  1. A.Improving the model's overall prediction accuracy
  2. B.Reconstructing training data features from model outputs✓ Correct
  3. C.Making the model inference significantly faster
  4. D.Augmenting the training data with synthetic samples

Explanation

Model inversion attacks exploit model predictions to reconstruct sensitive training data features, potentially exposing private information like faces or medical records.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q24. What is algorithmic recourse?

  1. A.Providing individuals with actionable steps to change an unfavorable AI decision✓ Correct
  2. B.Completely removing the algorithm from production deployment
  3. C.Using a completely different dataset for model retraining
  4. D.Making the algorithm run faster through hardware optimization

Explanation

Algorithmic recourse gives individuals affected by automated decisions information about what changes they could make to receive a different (favorable) outcome.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q25. What is the concept of Fairness through Awareness?

  1. A.Completely ignoring all protected attributes during model prediction
  2. B.Making random predictions without considering any input features
  3. C.Using the exact same model configuration for every single individual
  4. D.Treating similar individuals similarly using a task-specific similarity metric✓ Correct

Explanation

Fairness through Awareness formalizes individual fairness by defining a task-specific similarity metric, requiring that similar individuals receive similar predictions.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q26. What is the difference between equality and equity in AI fairness?

  1. A.Equity ignores all differences between individuals and groups
  2. B.Equality is always strictly the better approach for fairness goals
  3. C.They are completely identical concepts with no meaningful differences
  4. D.Equality gives the same treatment; equity adjusts treatment for fair outcomes✓ Correct

Explanation

Equality means identical treatment for all groups. Equity recognizes that different groups may need different treatment to achieve fair outcomes, accounting for historical disparities.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q27. What is the EU AI Act?

  1. A.A supervised machine learning algorithm for classification tasks
  2. B.A structured data format for storing model configurations
  3. C.A risk-based regulatory framework for AI in the European Union✓ Correct
  4. D.A general-purpose compiled programming language specification

Explanation

The EU AI Act is a comprehensive regulation that categorizes AI systems by risk level (unacceptable, high, limited, minimal) and imposes requirements proportional to the risk posed.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q28. What is homomorphic encryption and how does it relate to AI privacy?

  1. A.A compression algorithm for reducing data file sizes
  2. B.A process of permanently deleting data from storage
  3. C.Encryption allowing computation on encrypted data without decrypting it✓ Correct
  4. D.A standard symmetric encryption method for securing files

Explanation

Homomorphic encryption enables computation directly on encrypted data, allowing ML models to process sensitive data without ever decrypting it, providing strong privacy guarantees.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q29. What is the concept of AI safety?

  1. A.Making AI systems process data at significantly faster speeds for better user experience
  2. B.Ensuring AI systems behave as intended without causing unintended harm as capabilities grow✓ Correct
  3. C.Marketing and promoting AI products and services to potential commercial customers
  4. D.Reducing the overall financial costs of training and deploying large AI model systems

Explanation

AI safety is a research field focused on ensuring that AI systems are robust, reliable, and do not cause unintended harm, particularly important as AI systems become more capable and autonomous.

Report an error in this question

AI Ethics, Security & PrivacyHard

Q30. What is the difference between disparate treatment and disparate impact?

  1. A.Disparate treatment is the unintentional adverse effect on specific groups
  2. B.Disparate treatment uses protected attributes intentionally; disparate impact causes unequal effects
  3. C.Disparate impact is the intentional use of protected attributes in a decision model✓ Correct
  4. D.They are completely identical legal concepts with no meaningful differences

Explanation

Disparate treatment explicitly uses protected attributes in decisions. Disparate impact occurs when a seemingly neutral practice disproportionately affects a protected group, even without explicit use of protected attributes.

Report an error in this question

More AI / Machine Learning & Data Analytics Topics

Introduction to AI, ML & Data Analytics

Practice MCQs with explanations

Mathematical Foundations

Practice MCQs with explanations

Python for AI

Practice MCQs with explanations

Data Collection & Pre-processing

Practice MCQs with explanations

Exploratory Data Analysis

Practice MCQs with explanations

Supervised Learning

Practice MCQs with explanations

Ready to test yourself on AI Ethics, Security & Privacy?

Take a timed quiz drawn from 210+ questions on this topic. No signup required — your progress saves in your browser.

Start AI Ethics, Security & Privacy Quiz