By: Muhammad Faizan Khan
We thought data was just numbers and clicks, we think we’re using Artificial Intelligence. But mostly, it’s using us.
You might think you’re just browsing. Just scrolling. Just searching. However, from the moment you wake up and reach for your phone, the data collection begins. Location. App usage. Fingerprint pressure on the screen. Even how long you hover over a post before moving on. Every digital move we make, from what we search to what we scroll past, feeds an invisible system that is learning who we are. Not just our names or birthdays but our patterns, our impulses, our beliefs.
That’s not paranoia. That’s the design.
We live in a system where artificial intelligence doesn’t just observe us. It interprets us. And what it builds from that interpretation is a model so detailed, so persistent, it eventually becomes a stand-in for who we are. This isn’t about personalization anymore. It’s about identity extraction.
From Metadata to Mirror
Let’s start with a fact that still surprises people. Your data isn’t valuable because of what it says. It’s valuable because of what it reveals.
A timestamped purchase. A retweet. A Google search at 2:13 a.m. These fragments seem trivial. But together, they form a behavioral fingerprint. One that’s deeply difficult to fake and nearly impossible to erase.
By 2024, researchers at Northeastern University demonstrated that just seven days of app usage data could predict someone's personality traits with over 80 percent accuracy. Not just broad strokes, either. Down to traits like neuroticism or openness which influence everything from how you vote to how you spend money.
That same year, an insurance company in Japan used AI models trained on smartphone sensor data to estimate clients’ risk tolerance, adjusting policy offers based on movement patterns, not disclosures. People were being redefined by their data trails in real time without people even realizing it.
Identity Without Consent
This raises a deeper question. If AI decides your preferences, your risk level, your reliability, your creditworthiness or your psychological profile based on data, where is your input in that process?
Here’s the thing. We used to believe identity was something we constructed. Through reflection. Through experience. Through relationships. Now, much of that process is being automated.
Take hiring. More and more employers use AI-based personality traits through things like evaluations based on factors such as voice intonation, sentence structure or even micro-expressions using video interviews. You may feel you are presenting your best self. But the algorithm is scanning for patterns that align with its training data. The person you believe yourself to be is no longer the final authority.
What this really means is, AI doesn’t need to understand you in the way a friend does. It only needs to recognize you as a pattern.
The Efficiency Trap
This system is frighteningly efficient. But efficient for whom?
For advertisers, this indicates hyper-targeted ads designed to introject information that’ll reach the right person at the right time when they’re most likely to average faster than someone else making decisions. For governments, it means faster sorting, flagging and classification. For tech platforms, it means longer engagement, more predictive power and less friction.
But for you? It means your identity is being flattened. Reduced to vectors and weights inside a neural network. It means your complexity is treated as a computational problem to be optimized for someone else's goal.
And once an AI model builds its version of you, good luck challenging it. These systems aren’t required to explain themselves. There’s no appeals process for a decision made by an opaque algorithm trained on 500 million data points from people who vaguely resemble you.
When Prediction Becomes Personality
Here’s the uncomfortable truth, AI doesn’t need to understand you to define you. It just needs to predict you accurately enough to sell your attention or assess your value. That means your digital identity isn’t based on who you are but on how useful your behavior is to the system’s goals. If the algorithm decides your likely future involves debt, you might see more ads for payday loans. If it detects anxiety, you might be shown soothing content that keeps you scrolling, not necessarily healing.
You become the product of prediction. And when enough of these algorithmic predictions accumulate, they start to loop back into your real life, your moods, your decisions, even your sense of self. It’s one thing to see yourself through a mirror. It’s another to live inside one.
That same year, an insurance company in Japan used AI models trained on smartphone sensor data to estimate clients’ risk tolerance, adjusting policy offers based on movement patterns, not disclosures. People were being redefined by their data trails in real time, without even realizing it.
Redefining What It Means to Be Known
We used to think of being known as something emotional, friends who got us, partners who understood us. Now, to be known means being modeled, analyzed and statistically rendered. It’s an intimacy without empathy. AI doesn’t care if it misrepresents you, as long as the prediction holds. That’s not knowledge, it’s control disguised as understanding.
And yet, there’s a paradox here, many people crave the same thing the machines promise, recognition, validation, a sense of being seen. AI gives that illusion at scale. But it’s not seeing you. It’s seeing the data-trail of your decisions, the residue of your impulses, the measurable parts of your humanity.
The False Binary of Privacy
It’s tempting to frame all this as a privacy issue. But the bigger problem isn’t just who sees your data. It’s what gets constructed from it.
A digital identity that shapes your choices before you even make them.
A model that decides which jobs you’re shown. Which articles appear in your feed. Which political candidates are algorithmically aligned with your profile.
And here’s the brutal truth: opting out of data collection doesn't reclaim your identity. The model has already been trained. You’re already inside it.
Where Do We Go From Here?
We need to stop talking about data as just something we give away. It is not just metadata. It is identity. And identity, when abstracted and monetized, becomes something else entirely. It becomes a commodity.
There are efforts to bring the power back to users. The EU has implemented the Digital Services Act to allow people greater control over their spends of time and how they are profiled online. Companies (like Apple) are using privacy as a feature, not as an afterthought. These are directionally correct efforts but they do not fundamentally solve the problem.
We need a cultural shift in how we understand the relationship between self and system. AI doesn’t just use data. It interprets it. It builds a version of you, not from your words but from your behavior. And that version is increasingly the one that matters most in how the world interacts with you.
This isn't science fiction. It’s happening now.
So the next time you’re online, think about who’s watching. Not in a paranoid way but in a real one. The watcher isn’t a person. It’s a mirror. And it never blinks.
