AI & Digital Ethics Grade 6 AI Ethics Privacy Consent Accountability

Grade 6 | AI & Digital Ethics
Topic: AI Ethics – Privacy, Consent, Accountability

1. The Driving Question

"If a robot at the mall scans your face to recommend a smoothie, does it need your permission? And if it messes up—like suggesting a strawberry smoothie to someone allergic to strawberries—who’s in trouble: the robot, the mall, or the person who programmed it? How do we decide what’s fair when machines make decisions about us?"

2. The Core Idea – Built, Not Listed

Imagine you’re at Sunnyvale Mall, and a new kiosk called "SmoothieBot" greets you. It has a camera that scans your face, checks your age (to see if you’re old enough for the "Energy Boost" smoothie), and remembers your last order. The kiosk’s screen says, "Welcome back! Based on your past orders, we recommend the Mango Tango—would you like to try it?"

Here’s the puzzle: You never signed up for this. The mall didn’t ask if you wanted your face scanned, and you didn’t click "I agree" to anything. But now SmoothieBot knows your face, your age, and your smoothie habits. If it shares that info with the pizza place next door—or worse, if it misidentifies you as someone else and gives you the wrong smoothie—who’s responsible?

This is where AI ethics comes in. It’s not just about whether AI can do something (like scan faces), but whether it should—and what happens when things go wrong. Three big ideas help us answer that:

• Privacy: Your right to control who knows things about you. Example: If your school’s AI attendance system tracks your location all day, that’s different from a teacher taking roll call with a clipboard.
• Consent: Giving clear, informed permission for something to happen. Example: When you download a game, the "Terms & Conditions" pop-up is asking for consent—but most people click "Agree" without reading. Is that real consent?
• Accountability: Figuring out who’s responsible when AI causes harm. Example: If a self-driving car crashes, is it the car’s fault, the company that made it, or the person who was "supervising" it?

Grade 6 Note: In high school, you’ll learn that consent gets even trickier with AI—like whether a chatbot can "consent" to share your data with advertisers. In college, accountability becomes a legal debate: Can a company hide behind "the AI did it" if their algorithm discriminates?

3. Assessment Translation

How This Appears on State Tests (or Classroom Assessments):
- Multiple Choice: Questions will test whether you can identify privacy risks or consent issues in a scenario. Distractor patterns: - Confusing privacy with security (e.g., "The AI was hacked" vs. "The AI collected data without permission").
- Assuming consent is always given (e.g., "The user clicked ‘Agree’" when the terms were unclear).
- Blaming the AI itself for accountability (e.g., "The robot made a mistake" instead of "The company didn’t test it properly").
- Short Answer: You’ll be given a scenario (like SmoothieBot) and asked to: - Explain one ethical concern (privacy/consent/accountability).
- Suggest one way to fix it.
- Proficient response: Uses specific details from the scenario and vocabulary terms correctly.

Model Proficient Response (Short Answer):
Prompt: "A school uses an AI app called ‘Homework Helper’ that scans students’ handwriting to check their work. The app stores the scans to ‘improve its accuracy.’ What’s one ethical concern here, and how could the school fix it?"

Response: One concern is privacy—the app is collecting students’ handwriting without them knowing how it’s used or who can see it. The school could fix this by making sure students and parents consent before the app is used, like sending home a permission slip that explains what data is collected and how it’s stored. They should also let students opt out if they don’t want their handwriting scanned.

What Makes This Proficient: ✔ Names the specific concern (privacy).
✔ Explains why it’s a problem (data collection without transparency).
✔ Suggests a realistic fix (consent + opt-out).
✔ Uses vocabulary terms correctly.

4. Mistake Taxonomy

Mistake 1: The "It’s Just a Machine" Excuse
Prompt: "A social media app’s AI recommends videos to users. Some videos make false claims about vaccines. Who is accountable for the harm caused by these videos—the AI, the app’s developers, or the users who share them? Explain." Common Wrong Response: "The AI is accountable because it chose the videos." Why It Loses Credit: - Misunderstands accountability: AI doesn’t have legal responsibility—people do.
- Ignores the prompt’s "explain": Doesn’t say why the developers/users are responsible.
Correct Approach: The developers are accountable because they designed the AI to recommend videos, and they didn’t add safeguards to stop false claims. Users who share the videos also share blame, but the app’s creators have more power to fix the problem (e.g., by fact-checking videos before recommending them).

Mistake 2: The "Clicking Agree = Consent" Trap
Prompt: "A 12-year-old downloads a game that asks for access to their contacts and location. The ‘Terms & Conditions’ say the game can share this data with advertisers. Did the child give consent? Why or why not?" Common Wrong Response: "Yes, because they clicked ‘Agree.’" Why It Loses Credit: - Misunderstands consent: Consent must be informed (the person understands what they’re agreeing to) and voluntary (they have a real choice).
- Ignores age: Kids can’t legally consent to complex data-sharing agreements.
Correct Approach: No, this isn’t real consent. The child likely didn’t read or understand the terms, and the game didn’t explain the risks in a way a 12-year-old would get. Also, kids can’t legally agree to data-sharing—parents should have to give permission.

Mistake 3: The "Privacy Is Just About Secrets" Error
Prompt: "A fitness app tracks your runs and shares your route with your friends. Is this a privacy issue? Explain." Common Wrong Response: "No, because it’s not a secret—you’re just running outside." Why It Loses Credit: - Narrows privacy: Privacy isn’t just about hiding things; it’s about control over your personal data.
- Misses the context: The app is sharing more than just your location (e.g., how fast you run, where you start/stop).
Correct Approach: Yes, this can be a privacy issue. Even if you’re running in public, the app is sharing patterns about you (like your routine or fitness level) without asking how you want that data used. Privacy means you get to decide who sees what, even if it’s not a "secret."

5. Connection Layer

1. Within Subject (AI Ethics) → Algorithmic Bias
   Understanding privacy helps you spot bias: If an AI is trained on data collected without consent (like facial scans of only one racial group), it might work poorly for other groups. Privacy violations can lead to unfair AI.

2. Across Subjects (AI Ethics → Civics)
   Consent in AI is like laws in government: Just like a law needs public debate before passing, AI systems should get input from the people they affect. Both are about balancing power and fairness.

3. Outside School → Targeted Ads
   Ever notice ads for shoes you just looked at? That’s AI using your data without explicit consent. Now you’ll see "Accept Cookies" pop-ups everywhere and realize: This is a privacy choice, not just an annoyance.

6. The Stretch Question

"If an AI assistant (like Siri or Alexa) records your conversations to ‘improve its responses,’ but you didn’t know it was recording, is that stealing? Why or why not?"

Pointer Toward the Answer: This isn’t just about legality—it’s about trust. Stealing usually means taking something without permission, but with AI, the lines blur: Is your voice "yours" if the company owns the device? Some argue it’s more like eavesdropping (listening without consent), while others say it’s just how tech works. The bigger question: Should companies have to ask for permission before using your data, even if it makes their AI "smarter"?