“Anonymity Is an Illusion”

Let me ask you something.

Have you ever wondered why you put up with all of this?

You’re not alone. Surveys consistently show that most consumers say they don’t trust companies that want to monitor their behavior. And yet — those same people keep using Google, keep using Facebook, keep the smart speaker plugged in on the kitchen counter.

Why?

One of the most common answers: at least the data is anonymous. They don’t know it’s me. I’m just one number among hundreds of millions.

That’s what we’re examining today.

Is that actually true?

“Anonymous” Doesn’t Mean What You Think

Let’s start with a study.

The standard line from Google and other platforms is that what they collect is metadata — aggregated across large numbers of users in ways that make it impossible to identify any specific individual.

Sounds reasonable.

But in the early 2000s, a researcher named Latanya Sweeney ran a study that quietly demolished this assumption. She found that with just three data points — date of birth, zip code, and sex — she could uniquely identify 87% of the US population.

Three data points. No name. No ID number. Nothing you’d think of as sensitive.

And those three data points exist in almost every database.

Zuboff cites legal scholar Paul Ohm, who calls these supposedly anonymous behavioral data stores “databases of ruin” — because once the data is re-identified, what’s inside can destroy a person. Re-identification, Ohm argues, makes all our secrets fundamentally easier to discover. Enemies can more easily connect us to facts that can be used to blackmail, harass, defame, or discriminate against us.

This isn’t theoretical. It’s been demonstrated repeatedly.

Location Data: You Think You’ve Disappeared Into the Crowd

Location data is another place where the illusion of anonymity falls apart.

Every person’s movement patterns are uniquely their own. The route you take to work, the coffee shop you visit on Saturday mornings, how often you go to the pharmacy, when and where you see a doctor — together, these form what researchers call a “mobility signature.”

Even if the dataset contains no names, even if your phone number has been stripped out — with enough location data, any analyst with the right tools can pull you out of the crowd and identify you specifically.

In 2013, researchers from MIT and Harvard demonstrated exactly this. The uniqueness of individual movement patterns makes it technically feasible to re-identify a specific person from a large anonymized location dataset.

Princeton computer scientists Arvind Narayanan and Edward Felten put it plainly: there is no known effective method to anonymize location data, and no evidence that it’s meaningfully achievable.

The promise of anonymized location data is, technically speaking, a promise that cannot be kept.

What Your Phone’s Sensors Are Saying

Beyond location, the sensors inside your phone are leaking more than most people realize.

Accelerometer, gyroscope, magnetometer — these were designed to tell your phone which way it’s oriented, to help maps track your movement. They seem harmless.

But researchers have found these sensors can reveal far more. The slight change in your gait today — maybe you’re injured, maybe you’ve been drinking. The unusual stillness of your phone — maybe you’re ill. The pattern of vibration — your mode of transport, your emotional state.

Studies have shown that smartphone sensor data alone can be used to infer an ever-growing range of human activities and moods, and to extract sensitive personal information about specific individuals from anonymized datasets.

These sensors can be accessed by apps without explicit permission. Many apps collect this data continuously in the background, invisibly.

Facebook Knows Your Personality Better Than Your Friends Do

I want to look at one more study, because it pushes the anonymity question into new territory.

A team at Cambridge University led by researcher Michal Kosinski studied Facebook “likes.” They found that by analyzing what a person likes, they could predict personality traits with striking accuracy — introvert or extrovert, emotionally stable or neurotic, open or conventional.

A 2015 paper showed that their algorithm’s assessments of personality outperformed those of the subject’s coworkers, friends, and even family members.

With 300 likes, the algorithm knows you better than your spouse does.

The research was academic. But what it demonstrated is worth taking seriously: the seemingly casual traces you leave online — what you engage with, how long you pause, what you share — together, they build a portrait of you that may be more accurate than the one you’d draw of yourself.

And that portrait can be bought and sold.

It’s worth noting that this research quickly attracted funding from Microsoft, Boeing, Google, the National Science Foundation, and DARPA. The distance between academic research and commercial or military application is shorter than most people assume.

The Problem With “Consent”

This brings me to something I think is fundamental — the question of consent.

Every time we install an app, we click “agree.” We consent to terms of service that no one reads. The platforms say: you made a free choice.

Zuboff challenges this at its foundation.

She argues that genuine consent requires two things: first, that you actually understand what you’re agreeing to; second, that you have a real ability to refuse.

Neither condition exists in today’s digital world.

No one truly understands what they’re agreeing to — the terms are written in a complexity that would take even a lawyer considerable time to parse. And refusing isn’t a simple opt-out. If you refuse certain apps, you don’t just lose a service — you lose access to modern social life. Refuse certain messaging platforms and you lose your primary way of communicating with most people you know. Refuse maps and navigating daily life becomes genuinely difficult.

This is not a free choice. It is a designed absence of choice.

What I Keep Thinking About

Reading through this section, one thought stayed with me.

Most of us accept data collection because we believe we’re just anonymous numbers in a vast pool. That belief gives us a feeling of safety — they don’t know it’s me.

But this book shows that feeling of safety is itself something the system relies on.

The promise of anonymity lowers our guard and makes the whole apparatus easier to accept. The technical reality is different: anonymity is temporary. Given enough computing power and enough data, re-identification is a matter of time.

And the deeper problem isn’t even about whether they know your name.

Even if they never learn your name, they can still shape your behavior, engineer your emotions, and guide your choices. Behavioral modification doesn’t need to know who you are. It only needs to know what type of person you are — and then design interventions targeted at that type.

Individual anonymity offers no protection against collective manipulation.

That’s Episode 3.

We’ve looked at the illusion of anonymity — three data points are enough to identify you, location data cannot be meaningfully anonymized, your phone’s sensors reveal more than you’d think, and behavioral data can build a portrait of you more accurate than your own self-description.

The system depends on one assumption: that you don’t know, or don’t care.

Next episode, we step back and look at a bigger picture — the same underlying logic, operating in different places, in the hands of different actors. What happens then?

This is Let’s Read Together. I’m Allen.

Last modified on 2026-06-04