The AI That Decides for You
The AI That Decides for You
When Your Phone Knows What You Want Before You Do
Last Tuesday, seventeen-year-old Maya unlocked her phone at 7:43 AM, and before she could even open her email, she saw a notification: "New episodes from shows you might like." The streaming service had somehow predicted that after a stressful math test, Maya would want to watch the same comfort show she'd been binging for three weeks. It was creepily accurate. But here's what Maya didn't know: that recommendation wasn't random. It wasn't a magic algorithm or a team of Hollywood executives studying her psychology. It was an artificial intelligence system making millions of micro-decisions per second—and somewhere in a corporate office, an entire team of people had spent months building, testing, and monitoring the rules that governed how that AI worked.
This is the invisible world you're living in right now. AI systems are making decisions that shape your life in ways both obvious and hidden. The videos you watch, the news you see, the prices you pay, whether your assignment gets flagged for potential plagiarism, even how quickly your parents' credit card gets approved—all of these involve AI systems making choices about you and your family. And here's the uncomfortable truth: most teenagers have no idea these systems exist, let alone that someone needs to make rules for them.
This chapter is about understanding why AI needs governance—why those rules matter—and discovering that you're already interacting with governed AI systems dozens of times every single day. By the end, you'll see the invisible scaffolding that keeps AI systems from going off the rails. You'll understand why billion-dollar companies employ entire teams just to manage their AI. And you'll be ready to start seeing AI governance not as some abstract corporate concern, but as something directly relevant to your life.
AI Systems Make Millions of Decisions Affecting People's Lives Every Day
Let's start with a reality check. How many decisions do you think an AI makes about you in a single day?
If you guessed dozens, you're severely underestimating. Try hundreds. Maybe thousands. Every time you scroll through social media, an AI decides which content to show you and which to bury. When you search for something online, an AI ranks billions of web pages to find what it thinks you want. When your school uses software to check your homework for plagiarism, an AI is making a judgment about your integrity. When your parents shop online, an AI determines what prices they see. When they use their credit card, an AI decides in real-time whether to approve the transaction or flag it as fraud.
These aren't minor decisions. They're decisions that affect what you believe, what you buy, what opportunities you get, and whether your family can access their own money.
The Recommendation Engine That Shapes Your World
Consider the AI recommendation systems that power every major social media platform and streaming service. You might think these are simply showing you content you might like. But the reality is far more complex—and far more influential.
When you open TikTok, the AI behind the "For You" page has approximately zero seconds to decide which video to show you next. It analyzes thousands of signals: videos you've watched, how long you lingered, whether you liked or shared, the time of day, your location, the device you're using, and even patterns from millions of other users with similar behaviors. It then serves you a video, watches your reaction in milliseconds, and adjusts its next recommendation accordingly.
The AI isn't just responding to what you want. It's sculpting your preferences in real-time. Studies have shown that recommendation algorithms can create "filter bubbles"—environments where users progressively see only content that aligns with their existing views and interests. If you watch a few fitness videos, the algorithm might feed you increasingly extreme workout content, then supplement it with supplement advertisements, then connect you to fitness communities, and before you know it, your entire digital world revolves around one narrow interest.
This isn't accidental, but it's also not necessarily malicious. The AI is optimizing for what keeps you engaged, because engaged users click on advertisements, and advertisements generate revenue. The system is working exactly as designed—but whether that design serves your best interests is a question that requires governance oversight.
The Plagiarism Detector That Judges Your Work
Here's a scenario that hits closer to home for most students. Your teacher runs your essay through a plagiarism checker—software that scans your work against billions of web pages, previous student submissions, and academic databases. The AI makes a judgment: original work or copied content. This judgment can affect your grade, your academic standing, and your reputation.
But here's what most students don't realize: these AI systems make mistakes. They flag false positives—correctly original work that gets incorrectly identified as plagiarized. They fail to detect sophisticated plagiarism where someone rephrases, paraphrases, or uses translation tools to disguise copied content. They struggle with common phrases, proper nouns, and technical terminology that might appear identically across multiple sources.
In 2020, a high school student in Massachusetts was accused of plagiarism when Turnitin—the most widely used plagiarism detection software in education—flagged her essay as potentially copied. The problem? She had written the essay about her grandmother's immigration story, and the AI detected "similarities" because many students had written similar immigration stories using common phrases and expected narrative structures. After weeks of stress and appeals, the charges were dropped, but the experience left a lasting impact on the student.
This is a governance failure. The school had implemented an AI system without proper understanding of its limitations, without clear procedures for challenging its decisions, and without human oversight to verify the machine's judgment. We'll explore how proper governance structures can prevent these situations throughout this book.
The Fraud Detection System That Decides If Your Family Can Access Their Money
Let's move to something even more consequential: the AI systems that protect (and sometimes inconvenience) your family's financial life.
When your parents use their credit card at an unfamiliar location or make a larger-than-normal purchase, an AI system makes a split-second decision: approve the transaction or block it as potentially fraudulent. This happens in roughly 50 to 250 milliseconds—faster than you can blink. The AI analyzes hundreds of variables: the merchant's location and category code, the time of transaction, the amount, your parents' typical spending patterns, whether they've traveled recently, and dozens of other signals.
This technology has saved consumers billions of dollars in fraud losses. But it also creates frustrating false positives. Imagine your family is on vacation in another country, trying to pay for dinner, and the AI blocks the card because the purchase looks unusual. Or imagine you're trying to buy concert tickets the moment they go on sale—high-demand, high-value—and the AI flags the transaction as suspicious.
These aren't hypothetical scenarios. They happen millions of times annually. The question becomes: who decides what signals trigger a block? Who determines how much inconvenience is acceptable to prevent fraud? Who reviews when the AI makes mistakes?
The answers require governance—the rules, processes, and human oversight that determine how these AI systems operate. Without governance, you're left at the mercy of algorithmic decisions that might or might not serve your interests.