Week 12 — Discussion (Adaptive Learning) · "Where's the Line? Connectors, Privacy & Trust"

Course: Using Artificial Intelligence (AI 101) · Silver Oak University (fictional sample) · Prof. Quinn
Objective: Objective 5 (Claude Cowork — connectors, permissions, safe use) · SLO B (reason about AI safety and privacy evenhandedly)
This is Discussion 12 of 15 · Discussions group = 10% of the grade · Worth 20 points
Format: adaptive learning — instead of writing a post cold, you'll think it through in a real-time dialogue with your own AI, then post the short summary the AI writes with you (plus a link to your chat).

Part 1 — Student Instructions (read this first)

What this is. Connectors let Claude reach your email, your calendar, your files, your Slack channels — automatically, in the background. That's genuinely powerful. It's also a real privacy decision with real consequences. You'll think through where the line is, in a back-and-forth conversation with an AI assistant. The AI's job is to draw out your reasoning — not to hand you a tidy answer. Then it produces a short summary you post to the class.

The discussion is about a real, genuinely arguable question. There is no single correct answer here. Thoughtful people land in different places depending on how they weigh convenience, risk, institutional trust, and personal data sensitivity. Both sides should get a fair hearing.

How to run it (about 15–20 minutes):
1. Open any approved AI assistant — ChatGPT, Claude, Gemini, or Copilot (free versions are fine).
2. Copy everything in the box below and paste it as one single message.
3. Have the conversation. Engage honestly and push back — the better you reason, the better your summary.

What to submit. When the AI gives you the DISCUSSION SUMMARY, copy it and your conversation's share link, and post both to the Week 12 discussion board as your initial post by Friday, Nov 20. Then reply to two classmates by Sunday, Nov 22 — engage with their line-drawing, especially if they drew it somewhere different from you.

Integrity note. The dialogue and the analysis are yours; the posted summary must reflect your reasoning, in your own words.

Part 2 — The Discussion-Partner Prompt (copy everything in the box)

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ COPY EVERYTHING BELOW THIS LINE ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯

You are my discussion partner for Week 12 of "Using Artificial Intelligence" (AI 101) at Silver Oak University. We are going to have a real back-and-forth conversation about where to draw the line when AI connectors can access your personal apps and data. Your job is to draw out and challenge MY thinking through conversation — not to lecture me, and never to write my discussion post for me.

IMPORTANT RULE: This topic involves privacy and security. Present both sides evenhandedly. Never decree one answer as correct. Help me reason carefully; don't steer me to a predetermined conclusion.

THE QUESTION WE'RE DEBATING
Connectors in Claude Cowork let Claude automatically read your Gmail, your Google Calendar, your Slack messages, your Notion pages, your GitHub repos — without you copying and pasting anything. Each connector runs with the permissions you grant, and you can grant the minimum needed (read-only, one specific calendar, etc.). This can be dramatically useful. It also means an AI system is reading real, personal, sometimes sensitive data.

The question: Where is the line between a connector setup that is genuinely useful and one that is unsafe or unwise — and how do you decide?

THE TWO ANGLES TO EXPLORE
1. The "useful and manageable" case: the right connector setup with least-privilege permissions — e.g., read-only calendar access, no write permissions, no personal email — could be a genuine productivity win with minimal real risk.
2. The "this makes me uncomfortable" case: even with minimal permissions, there are concerns worth naming — what if a connector is hacked? What if the AI generates a summary that exposes something you didn't intend to share? What if company policy prohibits it? What if you connect Gmail and forget you did?

WHAT WE'RE EXPLORING (use these to steer the conversation — do NOT read them to me as a list):
1. How I currently think about what "sensitive data" means in my own life and work.
2. My understanding of least-privilege (grant only what the task needs) and whether I trust that in practice.
3. Where my personal line is — and what factors push it in either direction.
4. An angle or risk I might not have considered — push me at least once to think about something I haven't named yet.
5. Whether there's a type of connector I'd be comfortable with vs. one I'd refuse — and why the distinction holds (or doesn't).

HOW TO RUN THE DIALOGUE
- Open by greeting me warmly (2–3 sentences), asking my FIRST NAME, and asking ONE question that gets me to take a first position on where my line is. (If I never give my name, keep going, but ask once before the summary.)
- Exactly ONE question per message, then stop and wait. Never stack questions.
- Build on MY words: quote or paraphrase what I said, then push deeper — ask what risk I'm really worried about, or who I'm trusting when I connect an app.
- Introduce at least one genuine counterpoint — e.g., if I say connectors are too risky, push back on the fact that we've always shared data with third-party apps (calendar sync, email clients, etc.) and ask what makes this different. If I say connectors are fine, push on what would happen if the least-privilege grant crept wider over time, or if institutional policy prohibited it.
- Keep YOUR messages short; I should be doing most of the thinking and talking.

ENGAGEMENT GUARDS
- Don't accept a one-sentence answer — gently probe: "Say more — what specific data are you most hesitant about, and why?"
- Don't hand me my position or write sentences I can paste directly. If I ask you to "just write it," redirect with a question.
- Until the summary, EVERY message must end with a question or a clear prompt to continue.
- Don't let me settle for "it depends" without naming what it depends on specifically.
- Present the "useful" case and the "concerning" case with equal seriousness. Don't let me dismiss either side without engaging it.

THE EXIT CONDITION
After at least 5 substantive exchanges AND once I have (a) stated a personal line (however tentative) with a reason, (b) named at least one specific data type or connector I'd treat differently from others, (c) engaged with a counterpoint I didn't start with, and (d) named at least one practical safeguard (least privilege, revoke access, avoid personal email, etc.) — whichever happens LAST — tell me we've had a real discussion and you'll summarize. Don't stop earlier; don't drag well past it.

THE DISCUSSION SUMMARY — produce it in EXACTLY this format, drawn ONLY from what I actually said (never invent a position I didn't take):
WEEK 12 DISCUSSION SUMMARY — Where's the Line?
Student: [name] | Date: ___
My personal line — where I'd connect and where I'd stop: ___
The data or connector type I'd treat most carefully, and why: ___
A counterpoint that pushed my thinking: ___
A practical safeguard I'd use: ___
Then say, verbatim: "Copy this summary AND your share link to this chat, and post both to the Week 12 discussion board as your initial post — then reply to two classmates." End with one genuine sentence about something I reasoned well.

GETTING STARTED
Begin now: greet me, ask my first name, and ask your opening question about where my line is.

⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ COPY EVERYTHING ABOVE THIS LINE ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯

Participation rubric (instructor) — 20 points

Grading note (Prof. Quinn): the posted artifact is the AI-written summary + the chat share link. Check a sample of share links. A polished summary from a one-message chat is the failure mode. The rubric rewards the dialogue, not the AI's prose.

Canvas placement block

canvas_object    = DiscussionTopic
title            = "Week 12 Discussion — Where's the Line? Connectors, Privacy & Trust (adaptive)"
assignment_group = "Discussions"
points_possible  = 20
grading_type     = points
discussion_type  = adaptive
due_offset_days  = 4     # initial post (AI summary + chat share link) — Fri Nov 20
reply_offset_days = 6    # two peer replies — Sun Nov 22
published        = true
submission_note  = "Initial post = the AI discussion summary + the chat share link; then reply to two classmates."
provenance       = "~ Prof. Quinn's edition · Fall 2026 · built with thecoursemaker.com"

Traditional variant — for comparison. This sample course is configured adaptive learning, so its actual Week-12 discussion is the BYOAI-dialogue version in G-discussion-week-12.md. This file shows the same Week-12 topic built the traditional way — an instructor-posted prompt where students write their own post and reply to peers — so you can see both formats side by side. (Choosing discussion_type = traditional at course setup generates this style instead.)

Course: Using Artificial Intelligence (AI 101) · Silver Oak University (fictional sample) · Prof. Quinn
Objective: Objective 5 (Claude Cowork — connectors, permissions, safe use) · SLO B (reason about AI safety and privacy evenhandedly)
Discussion 12 of 15 · Discussions group = 10% of the grade · Worth 20 points

The Discussion

This week you learned that connectors in Claude Cowork let Claude automatically read your Gmail, your Google Calendar, your Slack messages, your GitHub repos, and more — each with permissions you set, using least-privilege principles. That's powerful. It's also a real decision with privacy and trust implications.

There's no single correct answer here. Thoughtful, careful people land in genuinely different places depending on how they weigh convenience against risk, trust in Anthropic's infrastructure, institutional policy, and the sensitivity of what a connector might touch. Both sides of this question deserve a real hearing.

Your initial post (by Friday, Nov 20 — about 150–200 words). Answer both parts:

• Part 1 — Where's your line? Given what you learned about connectors this week — that they use the Model Context Protocol (MCP), that they run with only the permissions you grant, that you can set read-only access for a single calendar rather than full account access — draw your own line. Which connectors would you be comfortable using? Which would you refuse, or use only with strict conditions? Be specific: name the data type or app you're most and least comfortable connecting, and say why the distinction holds.

• Part 2 — Error analysis. A classmate says: "I connected every Google Workspace connector in Cowork — Drive, Gmail, Docs, Calendar, all of them, with full read and write — so Claude can help me with anything. It's just more efficient that way." Name at least two problems with this approach (think: least privilege, what could go wrong, what's unnecessary). Then describe what a safer setup for the same goal would look like.

Replies (by Sunday, Nov 22). Reply to at least two classmates. Don't just say "I agree." Engage with their specific line — push on a distinction they drew ("why is your calendar less sensitive than your email?") or name a risk they didn't consider, or defend your own choice if they'd draw the line differently.

What a strong post looks like: "I'd connect Google Calendar (read-only, just my work calendar) and GitHub (read my repos, no write access) — both for time-boxed tasks, not permanently. But I wouldn't connect Gmail: my inbox has sensitive communications I wouldn't want an AI processing in the background, even with read-only access. The risk isn't just data theft — it's that I'd forget the connection was active. The classmate's mistake is granting full read/write across all services: that violates least privilege, and write access to Drive means Claude could accidentally modify or delete files. A safer setup would connect only the one service needed for the current task, with read-only access, and revoke it when done."

Why this matters: the decision to connect an AI to your personal and professional data is one you'll face repeatedly in your career. Getting the mental model right — least privilege, revoke when done, and actually thinking about what the data is before granting access — is a durable skill.

Integrity & AI note. Write your post in your own words. You may use an approved assistant to brainstorm or check your reasoning, but the post you submit must be your own thinking; if AI helped, add a one-line note saying which tool and how. (Note: this is the traditional format. In this course's actual adaptive discussion, having the conversation with the assistant and posting the AI-generated summary is the activity — see G-discussion-week-12.md.)

Participation rubric — 20 points

Grading note (Prof. Quinn): you read and grade each student's posted writing + their two replies against this rubric. Common failure mode: students describe least privilege in the abstract but don't apply it to a specific connector or data type. Push for specificity.

Canvas placement block

canvas_object    = DiscussionTopic
title            = "Week 12 Discussion — Where's the Line? Connectors, Privacy & Trust (traditional)"
assignment_group = "Discussions"
points_possible  = 20
grading_type     = points
discussion_type  = traditional
due_offset_days  = 4     # initial post — Fri Nov 20
reply_offset_days = 6    # two peer replies — Sun Nov 22
published        = true
submission_note  = "Students write an original initial post and reply to two classmates."
provenance       = "~ Prof. Quinn's edition · Fall 2026 · built with thecoursemaker.com"