Ethical AI

A is correct: Fairness in AI means treating all people and groups equally—avoiding bias where the system discriminates based on protected characteristics like race, gender, or age.

B is wrong: Speed is about performance efficiency, not fairness.

C is wrong: Profit is a business goal, separate from ethical fairness concerns.

D is wrong: Removing human oversight actually increases the risk of unfair outcomes, rather than ensuring fairness.

A is correct. Privacy protects people's sensitive data (like health records or financial info) from being stolen, sold, or used without permission—which could cause real harm.

B is wrong: Privacy doesn't make AI faster; in fact, privacy protections might slightly slow systems down.

C is wrong: This contradicts privacy. Privacy actually *limits* what data AI can access, not enables it.

D is wrong: While more data can help AI learn, privacy is about protecting people, not boosting AI performance. These are separate concerns.

Correct Answer: A

Responsibility in AI means being answerable for what the system does—if an AI makes a harmful decision or causes damage, someone (developer, company, user) must be held accountable. This is about ethics and consequences, not capability.

Why the others are wrong:

• B (Making AI intelligent): Intelligence and responsibility are different things. You can have a smart AI that nobody is responsible for—that's actually dangerous.
• C (Operating without human help): Autonomy is a feature of AI systems, not a principle of responsibility. In fact, responsible AI often *requires* human oversight.
• D (Speed of processing): This is a technical performance goal, completely unrelated to ethical accountability.

A is correct: AI automation is already replacing some jobs while creating new roles, fundamentally shifting employment landscapes and required skills.

B is wrong: While AI *uses* energy (which could affect climate), AI itself doesn't directly cause global warming—that's from greenhouse gases.

C is wrong: AI actually increases internet use, not decreases it, since most AI systems operate online.

D is wrong: There's no evidence AI reduces human intelligence; in fact, it's a tool that can enhance how we work and learn.

A is correct: Fairness requires active effort—avoiding biased training data prevents AI from learning unfair patterns, and regular audits catch problems before they harm people.

B is wrong: Speed has nothing to do with fairness; a fast decision can still be unfair.

C is wrong: Learning from only one group guarantees bias, since the AI won't understand different people's needs fairly.

D is wrong: More complexity and data often make bias *worse*, not better, unless that data is carefully checked for fairness.

A is correct: AI systems learn from their training data, so testing on diverse datasets helps catch biases and ensures the system works fairly for all groups of people, not just majority populations.

B is wrong: More data doesn't guarantee fairness—if that data contains biases or comes from unreliable sources, you're just amplifying problems.

C is wrong: Cherry-picking only positive outcomes hides real-world biases and creates a false sense of fairness.

D is wrong: Ignoring past biases means they stay embedded in your AI system; you need to actively acknowledge and address them.

A is correct: This follows the "data minimization" principle—AI systems should only gather what's actually needed for their purpose. Less data collected = less risk if something goes wrong, and less exposure of your personal information.

B is wrong: Sharing all data with third parties is the opposite of privacy protection. You'd have no control over who sees your information.

C is wrong: Encryption shouldn't be optional—it should be standard practice for protecting sensitive data. Waiting for users to request it leaves data vulnerable in the meantime.

D is wrong: Keeping data "just in case" violates privacy principles. Data should be deleted once it's no longer needed, reducing unnecessary storage of your personal information.

Why A is correct:
Ethical AI use requires shared responsibility across three groups: developers must build safe systems with ethical safeguards, users must deploy AI responsibly, and regulatory bodies set standards and enforce compliance. It's a team effort.

Why the others are wrong:
- B (Only the AI itself): AI systems have no independent agency or moral compass—they're tools that reflect the choices made by people who build and use them.
- C (Internet browsers): Browsers are unrelated to AI ethics; they're just software for viewing websites.
- D (Only the users): Users can't fix flawed systems created by developers or establish industry standards—responsibility is distributed, not placed on one group alone.

A is correct: As AI automates tasks, workers in certain industries (like manufacturing, data entry, or customer service) may lose jobs faster than new opportunities emerge, causing economic hardship.

B is wrong: Learning new skills is actually a *positive* impact, not negative.

C is wrong: Reducing errors is beneficial, not harmful.

D is wrong: Helping doctors is a positive application of AI for society.

A is correct: Ethical considerations help developers identify potential harms (like bias or misuse) and design systems that work fairly for all users, not just some groups.

B is wrong: Ethics isn't about speed—in fact, thinking through ethical issues often *slows* development because it requires careful planning and testing.

C is wrong: Ethics isn't primarily a cost-saving measure. While some ethical practices might reduce future legal costs, that's not the main reason to consider ethics.

D is wrong: Transparency and public accountability are actually *part* of good AI ethics, not something to avoid. Keeping AI secret raises ethical concerns rather than addressing them.

A. Bias in the training data ✓
AI systems learn from their training data, so if that data reflects human prejudices or is unrepresentative, the AI will perpetuate unfair outcomes. This is the most direct threat to fairness.

B. The color scheme of the AI interface ✗
Visual design doesn't affect how the AI makes decisions or treats different groups fairly.

C. The speed of AI processing ✗
How fast an AI runs doesn't determine whether its decisions are fair or biased.

D. The cost of AI development ✗
Budget doesn't directly cause fairness problems—a cheap or expensive system can be equally biased or fair depending on its data and design.

Correct Answer: A

Privacy in AI is crucial because AI systems often collect and process huge amounts of personal data—like your location, preferences, and behaviors. Without strong privacy protections, this sensitive information could be misused, leaked, or sold, harming individuals.

Why the others are wrong:
- B (speed of development): Privacy actually *slows down* development because it requires careful safeguards, not the other way around.
- C (affordability): Privacy protections don't directly make AI cheaper; cost depends on computing power and resources instead.
- D (unsupervised function): Privacy and supervision are separate issues—privacy is about protecting data, not about whether humans oversee AI systems.

Why A is correct:
AI systems are tools created by humans—they don't have their own intentions or consciousness. The people who design, train, and deploy AI, plus those who choose how to use it, are the ones making decisions and should be held accountable for the outcomes.

Why the others are wrong:
- B: Computers are just hardware running code; they can't be responsible for anything.
- C: AI isn't autonomous—it follows instructions written by humans and produces outputs based on human-designed systems.
- D: Random internet users have no involvement in creating or controlling specific AI systems, so they can't be responsible for their decisions.

Correct Answer: A
AI systems learn from data, and if that data reflects historical discrimination or bias, the AI will repeat and amplify those unfair patterns—for example, denying loans or jobs to certain groups. This is a genuine societal harm.

Why the others are wrong:
- B (more video games): Creating entertainment isn't a negative impact.
- C (making school easier): Easier school is beneficial, not harmful.
- D (helping with chores): Assistance with household tasks is a positive benefit.

A is correct: Ethical guidelines and regulations create frameworks that hold AI developers and users accountable, ensuring AI systems are transparent, fair, and don't cause harm.

B is wrong: Giving AI more freedom without oversight would actually increase risks of misuse and unintended harmful consequences.

C is wrong: Social media data is often biased, unreliable, and lacks quality—training AI on only this would make it *less* responsible, not more.

D is wrong: Secrecy prevents public scrutiny and accountability. Responsible AI requires openness so experts and society can assess and improve it.

A. Testing them with diverse data sets ✓

AI systems learn patterns from their training data. If you only test with limited data, the system might work well for some groups but fail for others—this is bias. Diverse datasets help catch these problems before deployment, ensuring fair treatment across different populations.

Why the others are wrong:

• B. Making them faster – Speed has nothing to do with fairness; a fast system can still be biased.
• C. Reducing complexity – Simpler isn't automatically fairer; you need the right *type* of testing, not just fewer features.
• D. Ignoring user feedback – Feedback helps identify fairness problems; ignoring it does the opposite.

A is correct: Privacy protections prevent unauthorized access to personal data that AI systems use for training. When user data is secure, people trust AI systems more, which encourages wider adoption and better datasets for development.

B is wrong: Privacy actually makes development *more* expensive—companies must invest in security, encryption, and compliance tools.

C is wrong: Privacy has nothing to do with gaming capabilities; it's about data protection, not game performance.

D is wrong: Privacy doesn't reduce the need for human oversight—in fact, responsible AI requires humans to monitor systems and ensure they use data ethically.

A is correct because responsible AI development requires thinking beyond just the technology itself—you need to consider how it affects real people and communities, including potential harms, fairness, and long-term consequences.

B is wrong because profit alone ignores ethical responsibilities and can lead to AI systems that harm users or society.

C is wrong because ethical guidelines exist specifically to prevent misuse and protect people; ignoring them is the opposite of responsible practice.

D is wrong because an AI system can be technically excellent but still cause real-world harm if it's biased, unsafe, or used unethically.

A. It helps build trust with users ✓
When AI systems are developed ethically—being fair, transparent, and accountable—people feel more confident using them. Trust is essential for AI adoption and long-term success.

B. It makes AI systems less accurate ✗
False. Ethical AI aims to be both fair *and* accurate. Ethics doesn't require sacrificing accuracy; it means ensuring accuracy works fairly for everyone.

C. It increases the complexity of AI systems ✗
Not necessarily. While ethical considerations may add some design steps, complexity isn't the main benefit or purpose of ethical AI.

D. It limits the use of AI ✗
Backwards logic. Ethical AI actually *expands* use by making systems more trustworthy and acceptable to users and regulators, not limiting it.

A is correct: Data protection laws exist specifically to protect user privacy. By following them, AI systems ensure they handle personal data legally and ethically—like getting consent, securing data, and letting users access their information.

B is wrong: Ignoring laws would do the opposite of protecting privacy—it's illegal and leaves users vulnerable.

C is wrong: Not respecting user rights is the definition of *violating* privacy, not protecting it.

D is wrong: Not updating policies means users don't know how their data is handled, which undermines privacy protection.

A is correct because AI ethics is fundamentally about creating trustworthy AI systems that treat people fairly, avoid discrimination, and make decisions responsibly—this is the core mission of the field.

B is wrong because power without ethics is dangerous; AI ethics actually constrains or directs power toward good outcomes rather than maximizing it.

C is wrong because AI ethics places safety as a priority equal to or above speed; rushing an unsafe AI system contradicts ethical principles.

D is wrong because a key part of AI ethics is *anticipating and addressing* unintended harms—ignoring them is the opposite of what ethicists do.

Why A is correct:
Encryption converts sensitive data into coded form that only authorized people can read, making it unreadable to hackers or unauthorized users—this is a fundamental privacy protection method.

Why the others are wrong:
- B (leaving data unprotected): This does the opposite of protecting privacy; unprotected data is easily accessed by anyone.
- C (ignoring encryption): Same problem as B—without encryption, sensitive data remains vulnerable.
- D (sharing encryption keys): Defeats the purpose of encryption; sharing keys with others allows them to decode the data, eliminating privacy protection.

Correct Answer: A

Both developers and users share responsibility because they create and control the AI's behavior. Developers design the system and are responsible for safety measures; users decide how to deploy it and must use it responsibly.

Why others are wrong:

• B (Only developers): Users make critical choices about *how* the AI is used, so they can't escape responsibility—a developer can't control misuse after release.
• C (AI itself): AI has no consciousness, intentions, or legal standing; it's a tool that does what it's programmed to do.
• D (No one): This creates a dangerous accountability gap where harmful outcomes go unaddressed and nobody has incentive to ensure safety.

A is correct: Fairness in AI means building systems that don't discriminate or disadvantage certain groups. It ensures decisions affecting people (like loans, hiring, or criminal justice) are unbiased and equitable.

B is wrong: Speed is about performance efficiency, not fairness—a fast AI can still be unfair.

C is wrong: This describes bias, the opposite of fairness. Fairness aims to prevent giving unfair advantages.

D is wrong: Cost reduction is an efficiency concern, not a fairness concern.

Why A is correct:
Transparency (showing how AI works) and accountability (holding developers responsible for outcomes) are the foundation of ethical AI. They allow researchers, users, and regulators to identify and fix problems, building public trust.

Why the others are wrong:

• B (keeping AI secret): Secrecy hides problems and prevents oversight—the opposite of ethics. It enables misuse without detection.

• C (focusing solely on profitability): Profit-only goals ignore potential harms to users and society. Ethical development requires balancing business with responsible practices.

• D (avoiding user feedback): Feedback helps developers understand real-world impacts and improve systems. Ignoring users means missing ethical concerns and potential harm.

A is correct: This describes bias — when an AI system makes unfair decisions based on protected characteristics like race, gender, or age. This is a genuine ethical problem that can cause real harm.

B is wrong: Processing speed is a technical limitation, not an unfairness issue. It affects efficiency, not fairness.

C is wrong: High cost is an accessibility problem, not an unfairness problem. Unfairness is about discriminatory treatment, not price.

D is wrong: A simple interface is actually *good* design — it makes the system easier for everyone to use. This has nothing to do with fairness or bias.

A is correct: Privacy is a fundamental safeguard in AI because these systems often collect and process large amounts of personal data. Privacy protections ensure that sensitive information about users isn't misused, exposed, or exploited.

Why others are wrong:
- B: Privacy doesn't inherently make AI more complex—it's a separate concern about data protection, not a technical complexity issue.
- C: Privacy actually *slows down* development because it requires careful data handling, consent processes, and security measures.
- D: Privacy doesn't reduce data storage needs; companies still store data, they just protect it better and use it more responsibly.

A is correct: AI developers have a core ethical responsibility to build systems that are safe and don't discriminate against or hurt people or communities. This involves testing for bias, security flaws, and harmful outputs.

B is wrong: Visual design (colors, appearance) is nice-to-have but not a fundamental developer responsibility—safety and fairness come first.

C and D are wrong: These are features that might be programmed into specific AI applications, but they're not universal responsibilities of AI developers. Safety and preventing harm apply to *all* AI systems.

Why A is correct:
AI automates tasks and creates new jobs while eliminating others, genuinely reshaping industries and employment. This is a well-documented, real effect already happening.

Why the others are wrong:
- B (taller): AI has no biological mechanism to affect human height—this isn't science-based.
- C (turning humans into robots): This is science fiction, not a realistic outcome of AI technology.
- D (weather patterns): While AI might help us *understand* or respond to climate change, AI itself doesn't directly control weather systems.

A is correct because critical thinking about AI helps us anticipate problems (bias, job loss, privacy issues) and design solutions so the technology helps society broadly rather than just a few people or causes damage.

B is wrong — we want AI to be *useful and accessible*, not harder to use.

C is wrong — making AI expensive doesn't address safety or fairness concerns; it just limits who can benefit.

D is wrong — AI isn't secret and can't stay secret; critical discussion actually helps society use it responsibly.