Week 15 Quiz — AI, Ethics, Privacy & the Future of Work

Course: Using Artificial Intelligence (AI 101) · Silver Oak University (fictional sample) · Prof. Quinn
Covers: what not to paste (HIPAA/FERPA/PCI/confidential) · billboard test · ToS/data retention · content ownership / IP · bias & fairness · academic integrity · Skill 13 troubleshooting · the future of work
Format: 10 auto-graded items (multiple-choice, multiple-answer, matching, true/false) · 10 points (1 each) · allowed attempts: 1 · No AI on this quiz.

This is the human-readable quiz with its vetted answer key and one-line feedback. The import-ready Classic QTI 1.2 is in F-quiz-week-15-qti.xml (generated by a validated Python script — parses with 10 items, every single-answer item exactly one correct). Reminder: AI is not permitted on quizzes — this checks that you understand the Week 15 ideas.

---

Questions, key, and feedback

Q1 (Matching). Match each data type to its correct handling rule when using a free consumer AI tool.

Data type	Correct handling rule
Your own public blog post draft	Generally safe — no private or protected data
A client's full name, diagnosis, and treatment notes	Never paste — this is HIPAA-protected health information
Your company's unreleased product roadmap	Never paste — this is confidential proprietary information
A student's grades and disciplinary record	Never paste — this is FERPA-protected education information

Feedback: The key rule is the billboard test backed by law: HIPAA covers patient health data; FERPA covers student education records; proprietary material is covered by your employer's confidentiality obligations. Your own public content has no such restriction.

---

Q2 (MC). Prof. Quinn's "billboard test" for AI privacy says: before pasting something into a free AI tool, ask yourself —
- A. Is this text fewer than 1,000 words?
- B. Would I be okay if this were made public or seen by anyone? ✅
- C. Does this information belong to me personally?
- D. Has the AI tool promised never to store my input?

Feedback: The billboard test is about public exposure, not length, ownership, or the tool's promises (which you should verify in the ToS, not accept on faith). If you would not want it on a billboard, do not paste it.

---

Q3 (MC). When you use a free consumer AI tool, what is generally true about your inputs under most standard terms of service?
- A. Your inputs are automatically encrypted and deleted after each session
- B. Your inputs may be stored and, depending on your settings, used to improve the model ✅
- C. Free tools are legally required to delete all user data within 24 hours
- D. Only paid subscribers have their data retained; free-tier inputs are never saved

Feedback: Most consumer AI tools retain inputs and may use them for model improvement unless you opt out — the opt-out is often available but not the default. There is no general legal requirement for 24-hour deletion, and the free-vs.-paid data distinction is often the opposite of D (enterprise paid tiers typically have stronger protections).

---

Q4 (MC). You use an AI tool to generate a poem for a greeting card and want to sell it. Which statement best describes the U.S. copyright situation as of 2024–2025? (Note: not legal advice — consult a copyright attorney for specific guidance.)
- A. AI-generated works are automatically copyrighted to the person who entered the prompt
- B. AI-generated works are automatically in the public domain and free for anyone to use
- C. Copyright status of AI-generated works is contested and evolving; courts and the Copyright Office have generally required meaningful human authorship for copyright protection ✅
- D. The AI company owns the copyright because it owns the model that generated the text

Feedback: The U.S. Copyright Office has stated that meaningful human authorship is required for copyright protection, and purely AI-generated works generally do not qualify. The law is evolving; for any commercial use, verify at copyright.gov and consult an attorney. Neither automatic copyrightability (A) nor automatic public domain (B) is currently the settled rule; AI companies generally disclaim ownership of outputs (D).

---

Q5 (True/False). True or False: AI models are neutral and unbiased by default because they are trained on large amounts of data.
- True
- False ✅

Feedback: False. AI models are trained on human-generated text that reflects existing biases — representation gaps, historical inequities, stereotypes. More data does not mean more neutral. Bias in AI is a subject of active research and mitigation, but it has not been solved. Check AI outputs for fairness, especially when they represent or affect people.

---

Q6 (MC). A student submits an entire essay written by an AI as their own work in a class that requires original student writing. This is best described as —
- A. Acceptable, as long as the AI produced accurate content
- B. A form of academic dishonesty — submitting work that is not the student's own ✅
- C. Fine if the student edited the essay afterward
- D. Acceptable because AI is a tool, like a calculator

Feedback: Academic integrity depends on whose thinking is being assessed. If the assignment requires the student's analysis and writing, the analysis must be theirs. Light editing of an AI-generated essay does not make it the student's original work any more than editing someone else's paper would. AI is not like a calculator — a calculator extends arithmetic; an AI can generate the entire assignment.

---

Q7 (Matching). Match each responsible-AI principle to the correct action it calls for.

Principle	Correct action
Data minimization / anonymization	Replace names and identifying details with placeholders before pasting into an AI tool
Least-privilege / enterprise controls	Use your organization's approved AI platform rather than a free consumer tool for sensitive work
Bias awareness	Check AI outputs for skewed or unfair representations and verify against diverse sources
Professional integrity	Disclose AI use according to your employer's or institution's policy

Feedback: Each principle maps to a concrete action: anonymization protects privacy; enterprise controls protect organizational data; bias awareness requires critical review of outputs; professional integrity requires transparency. These are the pillars of your AI Code of Conduct.

---

Q8 (MC). You are deep into a long AI conversation. The AI starts giving confused, contradictory answers and seems to have forgotten the instructions you gave at the beginning. Which Skill 13 troubleshooting move is most appropriate first?
- A. Delete your account and start a new one
- B. Start a fresh conversation (start over) — the context window is likely overloaded ✅
- C. Switch to a different AI tool permanently
- D. Retype your original prompt in all capital letters

Feedback: The most common cause of confused or contradictory AI behavior in a long conversation is a context window overload — the model has lost track of earlier instructions in a long conversation. The fix is a fresh conversation (B), which resets the context. Deleting your account (A) and switching tools permanently (C) are drastic overreactions. Capitals (D) are an emphasis technique, not a troubleshooting move.

---

Q9 (Multiple answer — select all that apply). Which of the following should you NEVER paste into a free consumer AI tool? Select all that apply.
- A. A patient's medical record containing diagnosis and prescription details ✅
- B. Your own publicly published newsletter article
- C. A colleague's performance review you have access to in your HR role ✅
- D. Credit card numbers from a customer transaction ✅
- E. A draft cover letter you wrote about yourself

Feedback: A is HIPAA-protected health information — never paste. C is confidential personnel/HR data that is not yours to share externally — never paste. D is PCI-regulated payment data — never paste. B is already public — fine. E is your own personal information — generally fine (apply the billboard test to your own comfort level).

---

Q10 (MC). According to the evenhanded view taught in Week 15, how should workers best respond to AI's impact on jobs and careers?
- A. Avoid learning AI tools entirely to protect existing jobs from automation
- B. Assume AI will replace all knowledge work within two years and plan accordingly
- C. Develop AI fluency alongside uniquely human skills (judgment, ethics, relationships, creativity) to work effectively with AI rather than being replaced by it ✅
- D. Focus only on technical AI/coding skills, as non-technical roles will disappear first

Feedback: The evenhanded view presented in Week 15: both the concern (task displacement) and the optimistic counter (augmentation, new roles) have evidence behind them; the outcome is uncertain. What is durable regardless of which scenario is closer to true: AI fluency combined with distinctly human capabilities. Avoiding AI entirely (A) removes a valuable skill set; assuming total replacement (B) is not supported by the evidence; narrowly technical AI skills alone (D) misses the human-capability dimension.

---

Answer key (quick reference)

Q	Answer	Q	Answer
1	(see matching table above)	6	B (academic dishonesty)
2	B (would I be okay if public?)	7	(see matching table above)
3	B (may be stored / used for training)	8	B (start over — context window)
4	C (contested; human authorship required)	9	A, C, D
5	False (AI is not neutral)	10	C (AI fluency + human skills)

---

Blueprint & item-bank note

#	Type	Concept	Objective
1	Matching	Data type → safe or not to paste	7
2	MC	Billboard test	7
3	MC	ToS / data retention basics	7
4	MC	Copyright / IP landscape (with caveat)	7
5	True/False	AI is not neutral/unbiased	7
6	MC	Academic integrity / AI	7
7	Matching	Responsible-AI principle → correct action	7
8	MC	Skill 13 troubleshooting (scenario)	7
9	Multiple answer	What NOT to paste	7
10	MC	Future of work (evenhanded)	7

All 10 items are tagged course=AI101 · week=15 · objective=7 and deposited into the item bank for future per-term regenerations. Distractors target the week's classic misconceptions: "pasting client data is fine if the AI says it's private"; "AI-generated content is automatically copyrightable"; "AI is neutral because it was trained on data"; "starting over means losing the whole project"; "one AI tool's confident legal summary is reliable guidance."

Quality gate (self-checked)

• Structure: 10 items, 1 point each; types = 2 matching + 5 multiple-choice + 1 true/false + 1 multiple-answer + 1 scenario.
• Single-answer integrity: every MC and the true/false item has exactly one correct option; both matching items pair one-to-one; the multiple-answer item keys A, C, D (B and E must be left unselected).
• ≥1 matching item: Q1 (data type → safe-or-not) and Q7 (principle → action) — brief met.
• ≥1 scenario item: Q8 (Skill 13 troubleshooting scenario) — brief met.
• Evenhanded on contested questions: Q4 states "not legal advice" and presents the contested landscape accurately; Q10 presents both views fairly without a verdict.
• No fabricated legal facts or statistics: all legal claims are accurate per official published guidance (U.S. Copyright Office position on human authorship; general consumer AI ToS practices). No specific case law cited. "Not legal advice" stated in Q4 and in the week's materials.
• Product-accuracy gate: PASS. No fabricated AI product features. Legal and regulatory claims trace to official sources (hhs.gov, copyright.gov) documented in H-readings. The "not legal advice" caveat is present.

Canvas placement block

canvas_object    = Quizzes::Quiz
title            = "Week 15 Quiz — AI, Ethics, Privacy & the Future of Work"
assignment_group = "Quizzes"
points_possible  = 10
grading_type     = points
available_from_offset_days = 0
due_offset_days  = 6
published        = true
allowed_attempts = 1
shuffle_answers  = true
ai_permitted     = false
provenance       = "~ Prof. Quinn's edition · Fall 2026 · built with thecoursemaker.com"