Chapter 4: Managing AI – Intentionality and Responsibility

Managing AI 🎛️

Intentionality and Responsibility

Now that we have explored how to engage with AI for information and collaborate with it for creation, it is easy to see the appeal of using these tools for absolutely everything. When you have a machine that can summarize a textbook chapter in three seconds, draft an email to your teacher, and solve a complex algebraic equation, the temptation is to put your entire brain on autopilot. AI can quickly feel like the ultimate “easy button” for life.

However, the third domain of the AILit Framework — Managing AI — requires us to take a step back. Managing AI is not about learning a new software trick; it is about developing the discipline to use technology intentionally. It involves recognizing that just because an AI can do something for you does not necessarily mean it should. Managing AI requires you to protect your personal data, recognize the inherent value of human struggle, and intentionally decide when to hand the steering wheel over to a machine and when to keep your own hands firmly on the wheel.

4.1 — The Economics of Free AI: You Are the Data 🔒

Most of the AI tools you interact with every day — from social media recommendation algorithms to powerful chatbots — are offered to you completely free of charge. But building, training, and running these massive AI systems costs billions of dollars in computer servers and electricity. This brings us to a fundamental rule of the modern internet:

If you are not paying for the product, you are the product.

When you use a free AI chatbot, your conversations do not just vanish when you close the tab. In many cases, the company behind the AI is collecting your prompts, your uploaded documents, and your feedback to further train and refine their next generation of models.

What’s at Risk?

If you ask an AI to:

📝 Rewrite a highly personal essay
🏥 Summarize a confidential medical document
💬 Help you vent about a conflict with a friend

…that private information is ingested by the system. While companies attempt to anonymize this data, privacy experts warn that pieces of sensitive information can sometimes resurface in the AI’s responses to other users.

Protecting Your digital footprint

Managing AI means being fiercely protective of your digital footprint. Before you type anything into a prompt box, operate under the assumption that a human engineer might read it or the machine might memorize it.

Responsible management of technology means:

📋 Taking the time to understand the terms of service
⚙️ Opting out of data sharing when the platform allows it
🚫 Establishing a strict personal boundary regarding the kind of personal information you are willing to trade for digital convenience

4.2 — Knowing When NOT to Use AI 🧠

The most difficult part of managing AI is knowing when to turn it off. In education and personal growth, the struggle to learn something new is not a bug — it is a vital feature.

When you wrestle with a difficult math problem, spend hours trying to articulate a complex thesis statement, or painstakingly learn to play a new chord on a guitar, your brain is building new neural pathways. This concept is known as productive struggle . If you use AI to bypass that struggle — by having it instantly generate the essay or solve the math equation — you get the final product, but you completely miss out on the cognitive workout.

It is the equivalent of paying someone else to lift weights for you at the gym and wondering why you aren’t getting any stronger.

The Danger of Automation Bias 🤖

Furthermore, we must be wary of automation bias . We have all heard stories of drivers blindly following their GPS navigation right into a lake because the machine told them to turn. The same phenomenon happens with AI.

If a spelling and grammar AI tells you to change a sentence, or a coding assistant suggests a block of code, the natural human instinct is to assume the computer is smarter than you and accept the change without thinking. Managing AI requires overriding this bias.

When Human Judgment Is Irreplaceable

There are specific domains where human empathy, nuance, and ethical judgment are irreplaceable. An AI can instantly draft a perfectly formatted apology letter to a friend you hurt, or a condolence message to someone who has lost a loved one. But delivering an AI-generated message in a moment that requires genuine human connection is ultimately hollow.

True emotional intelligence and ethical decision-making cannot be outsourced to an algorithm.

4.3 — The Future of Work and Human Augmentation 🚀

As AI continues to advance, the anxiety surrounding the “future of work” grows louder. The fear is that AI will replace human workers entirely, making traditional jobs obsolete. While it is true that AI will automate certain repetitive tasks, the reality is much more nuanced.

We are not moving toward a world where humans are replaced by machines. We are moving toward a world of human augmentation , where professionals use AI to enhance their own capabilities.

Profession	What AI Does	What the Human Still Does
Doctor	Rapidly analyzes thousands of medical images for early signs of disease	Delivers the diagnosis, comforts the patient, tailors the treatment plan
Lawyer	Summarizes thousands of pages of legal documents in seconds	Argues the case persuasively in front of a jury
Teacher	Identifies knowledge gaps from student assessment data	Builds relationships, adapts lessons to individual students, inspires curiosity
Journalist	Transcribes interviews and suggests headline options	Investigates, verifies, builds source trust, makes editorial judgments

The Ultimate Human Advantages

The professionals who thrive in the Algorithmic Age will be those who know how to manage AI as a powerful assistant while doubling down on their uniquely human skills:

🧩 Critical thinking and strategic planning
🗣️ Complex communication and storytelling
❤️ Emotional intelligence and empathy
🌱 Adaptable leadership in uncertain situations

Managing AI means positioning yourself not as an opponent competing against the machine, but as a skilled supervisor who directs the algorithm to handle the tedious work — freeing you up to do the deep, meaningful thinking that only a human mind can achieve.

4.4 — AI, Algorithms, and Teen Mental Health 🧠

Managing AI isn’t just about productivity and privacy — it’s about your wellbeing. The relationship between algorithmic platforms and adolescent mental health has become one of the most consequential public health questions of our generation.

The Research Landscape

Social psychologist Jonathan Haidt, in his 2024 book The Anxious Generation, documented a striking correlation: rates of teen depression, anxiety, self-harm, and loneliness began rising sharply around 2012 — precisely when smartphone adoption among teenagers became widespread and social media platforms shifted to algorithmically curated feeds. The correlation is not proof of causation, but it is striking across dozens of countries simultaneously.

More damning is the evidence that emerged when Facebook whistleblower Frances Haugen leaked internal company research in 2021. Facebook’s own internal slides stated, regarding Instagram: “We make body image issues worse for one in three teen girls.” The company knew. The research was internal. And the platform continued to deploy the same algorithmic features.

The dopamine loop

The mechanism is neurological. When you receive a like, a comment, or an unexpectedly funny video, your brain releases dopamine — the neurotransmitter associated with reward and motivation. This is the same circuit activated by food, social connection, and achievement.

Social media platforms engineer their products to trigger this dopamine release unpredictably — not every time you scroll, but sometimes, in an unpredictable pattern. As we discussed in Chapter 2, this is the variable reward schedule: the same psychological mechanism that makes slot machines so compulsively addictive. The platform doesn’t want you satisfied; it wants you perpetually seeking.

Passive vs. Active Use

Research by psychologists Amy Orben and Andrew Przybylski found an important nuance: not all screen time is created equal. Passive consumption — scrolling through content without interacting, watching others’ highlights without creating your own — is most strongly associated with negative mental health outcomes. Active creation — building something, commenting, genuinely connecting with specific people — has a much weaker or even neutral relationship with wellbeing.

This suggests a practical framework: when you use social platforms, notice whether you are creating and connecting (active) or silently consuming (passive). The algorithm will always push you toward passive consumption — it’s more scalable and more profitable. Choosing active engagement is an act of self-management.

4.5 — The Data Broker Industry 🕵️

Most people are aware that companies like Google and Meta collect data about them. Fewer people are aware of a vast, largely invisible industry that operates entirely in the background: the data broker industry.

Who Are Data Brokers?

Companies like Acxiom, Spokeo, LexisNexis, and hundreds of smaller firms compile detailed personal profiles on billions of people — without those people ever interacting with them or consenting to data collection. They aggregate data from:

Public records (property ownership, court filings, voter registrations)
Retail loyalty programs and purchase histories
Social media profiles and activity
Location data purchased from apps
Credit and financial records
Online browsing behavior

Your profile likely includes: estimated income bracket, political affiliation, health conditions (inferred from purchases), religious beliefs, relationship status, employment history, and a detailed location history showing where you spend your time.

Who Buys This Data?

The buyers are diverse and often surprising:

Marketers targeting ads based on your profile
Insurance companies pricing premiums based on inferred risk factors
Employers screening job applicants beyond what resumes reveal
Law enforcement agencies purchasing location data to track individuals without a warrant (a practice that several courts have found legally troubling)
Political campaigns micro-targeting voters with emotionally tailored messages

The surveillance economy

Harvard Business School professor Shoshana Zuboff coined the term surveillance capitalism to describe this system: an economy built not on making things, but on monitoring people. Your behavior — your clicks, your purchases, your locations, your social connections — is the raw material that fuels this industry.

Protecting Yourself

You cannot opt out of the surveillance economy entirely, but you can reduce your exposure:

Use opt-out portals: Major data brokers are required in some states to provide opt-out mechanisms. Services like DeleteMe or Privacy Bee automate this process.
Use a VPN: Prevents your internet service provider from selling your browsing history.
Review app permissions: Location access, microphone access, and contact access should be granted only when genuinely necessary.
Read privacy policies: Specifically look for the sections on data sharing with third parties. If the policy says your data can be shared with “affiliates” or “business partners,” that typically includes data brokers.

4.6 — Your Legal Rights Around Automated Decisions ⚖️

One of the most important things to understand about managing AI is what legal protections you currently have — and where the gaps are.

The European Union’s General Data Protection Regulation includes GDPR Article 22 , which grants EU residents:

The right not to be subject to decisions made purely by automated systems when those decisions significantly affect them (credit, employment, healthcare)
The right to request human review of any automated decision
The right to an explanation of how the automated decision was made

This is a powerful legal protection — but it applies only to residents of the European Union.

CCPA (California Consumer Privacy Act)

California residents have additional rights under the CCPA:

The right to know what personal data a company has collected about them
The right to request deletion of that data
The right to opt out of the sale of their data to third parties

Efforts to pass a federal U.S. equivalent have stalled repeatedly in Congress, leaving most Americans with no equivalent federal protection.

COPPA (Children’s Online Privacy Protection Act)

Federal law does protect children under 13 from commercial data collection without verifiable parental consent. This is why most platforms claim a minimum age of 13, and why school-deployed apps are subject to additional scrutiny. However, COPPA has significant enforcement gaps, and teenagers between 13 and 17 have virtually no federal legal protection comparable to EU children.

The Gap

The stark reality for most American teenagers: many of the AI systems that make consequential decisions about your life have no legal obligation to be transparent about how they work, no requirement to provide human review, and no pathway for you to challenge their outputs. This legal gap is one of the driving arguments for AI literacy — knowing how these systems work helps you navigate a world where the law hasn’t yet caught up.

4.7 — Building Digital Resilience 🏗️

Understanding the risks of AI and algorithmic systems is only half the battle. The other half is building the daily habits and mindsets that protect your autonomy, your attention, and your wellbeing. We call this digital resilience — not a rejection of technology, but an intentional relationship with it.

The 5 R’s Framework

1. Recognize

Learn to notice when you are inside an algorithmic loop. Signs include:

Scrolling without consciously choosing to continue
Feeling anxious or worse after a session, but continuing anyway
Losing track of time while consuming content you wouldn’t have chosen in the morning

The first step is simply noticing. Set a gentle phone alarm after 20 minutes of social media use — not to stop, but to pause and consciously ask: Do I want to keep going?

2. Reduce

Use the tools built into your devices:

Screen Time (iOS) / Digital Wellbeing (Android): Set daily limits for specific apps
Turn off autoplay: On YouTube, Netflix, and podcast apps — autoplay is designed to bypass your conscious decision to continue
Disable non-essential notifications: Every notification is a designed interruption engineered to pull you back to the platform

3. Reclaim

Schedule intentional offline activities that restore your attention span and dopamine regulation:

Physical exercise: Proven by research to have the strongest positive effect on teen mental health
Deep reading: Books, not articles — extended focus builds attentional capacity
Creative work without screens: Music, cooking, drawing, building — activities where you produce rather than consume
Face-to-face conversation: Particularly activities where phones are not present

4. Reflect

Once a week, ask yourself:

Was my AI and social media use this week intentional or reactive?
Did I learn something genuinely valuable from my time online?
How did I feel during and after my digital sessions?
What would I do differently next week?

A simple journaling habit — five minutes at the end of each week — builds the metacognitive awareness that makes all the other R’s more effective.

5. Resist

Actively and deliberately seek information from non-algorithmic sources:

Libraries and librarians: Human-curated knowledge without engagement optimization
Long-form journalism: The New Yorker, The Atlantic, local investigative journalism
Books: The original long-form information technology
Human experts: Teachers, mentors, community elders — people with accountability to their reputations

Every time you consult a human expert or a carefully edited book, you are choosing a different information ecosystem — one not optimized to maximize your time on platform.

Digital Resilience Self-Assessment

Score yourself honestly on each dimension (1 = Never, 2 = Sometimes, 3 = Usually, 4 = Almost Always):

Dimension	Score
I know when I’ve been passively scrolling for more than 20 minutes
I have app time limits set on at least two platforms
I regularly read books or long-form articles offline
I spend meaningful time in face-to-face conversation each day
I seek out news from multiple sources, not just my feed
I reflect on my digital habits at least once a week
I turn off autoplay on streaming platforms
I know how to opt out of data tracking in app settings

Total Score:

8–12: Building awareness — identifying habits to change
13–20: Developing resilience — practicing intentionality with room to grow
21–28: Strong digital resilience — modeling healthy habits for others
29–32: Exceptional — consider sharing your strategies with your community

Chapter 4 Project: The AI Policy Pitch 📋

The Challenge: You are going to step into a leadership role by drafting a clear, ethical policy that governs how AI can be used in a shared environment. This is not about banning technology — it is about creating a framework that balances the incredible productivity gains of AI with the need to protect human integrity, original thought, and privacy.

Step 1: Choose Your Environment

Pick one of the following settings (or invent your own) and tailor your language and rules to fit its specific needs:

🏫 High School English Department — An “Acceptable Use Policy” for essay writing
📰 School Journalism Club — A code of conduct for student reporters
🎨 Fictional Graphic Design Company — A corporate policy for professional designers
🤖 Competitive Robotics or Coding Team — Guidelines for using AI coding assistants

Step 2: Define the Boundaries

Write a one-page policy document that explicitly defines:

Policy Type	At Least…	Example
Encouraged Uses	2 specific allowances	”Writers are encouraged to use AI to brainstorm alternate titles”
Prohibited Uses	2 specific prohibitions	”Submitting an entirely AI-generated draft as your final essay is strictly forbidden”

Step 3: The “Why” (Rationale)

Rules without reasons are rarely followed. For every boundary you create, write a brief rationale explaining the ethical or practical reasoning behind it. You must reference key concepts from this chapter.

Step 4: The Pitch

Summarize your document in a short, persuasive paragraph designed to win the support of your audience. Convince them that these rules are designed to protect their unique human skills, ensure their work remains legally and ethically sound, and allow them to confidently use cutting-edge tools without fear of crossing an invisible line.

Step 5: Stakeholder Map

Before your policy can be implemented, you need to understand who it affects. Create a Stakeholder Map — a visual or written identification of every group touched by your policy and an explanation of how the policy protects (or constrains) their interests.

For example, if your policy governs AI use in a high school English department, your stakeholders might include: students, classroom teachers, department heads, school administrators, parents, college admissions officers, and future employers of your graduates. How does your policy serve each group’s legitimate interests?

Step 6: Implementation Timeline

Great policies fail without realistic implementation plans. Outline how your policy would be rolled out over one academic year:

Phase	Timeline	Action
Awareness	Month 1–2	Staff training; student orientation; FAQs distributed
Pilot	Month 3–5	Apply policy in 2–3 classes; collect feedback
Review	Month 6	Identify gaps and unintended consequences
Full Launch	Month 7–10	School-wide implementation with monitoring
Annual Review	Month 12	Assess effectiveness; update for new AI capabilities