Dear Algorithm, Who Am I? When AI Decides Your Type Before You Do

There’s a question I’ve been turning over for a while now, and it sits at the intersection of everything I’ve spent the last decade building. It’s not a marketing question, though it has profound implications for marketing. It’s not a technology question, though technology is what makes it urgent. It’s a question about identity, and it goes like this: when an AI system constructs a model of who you are, based on how you ask questions, what you click on, and how you respond to its recommendations, does that model change who you actually become?

I wrote a book called Dear Algorithm, It’s Not Me, It’s You, and the title was deliberately playful. But the underlying concern wasn’t playful at all. It was about the growing gap between who we are and who the algorithm thinks we are, and what happens when that gap closes, not because the algorithm got better at reading us, but because we got better at being read.

This is not a dystopian argument. I’m not suggesting that AI is eroding human agency or that we’re sleepwalking into algorithmic determinism. What I am suggesting is something more nuanced and, I think, more important. The feedback loop between AI-assigned identity and human behaviour is already operating. And the STAR Framework, built on seven established psychological theories, gives us a vocabulary for understanding exactly what’s happening inside that loop.

The Feedback Loop

Let me describe what’s actually happening, because the mechanism matters.

An AI assistant mediates your interaction with the world. You ask it questions. It responds. You ask follow-up questions. It adapts. Over time, the system builds a model of who you are. Not a formal psychometric profile. Not a labelled category. But a functional approximation that serves the same purpose: predicting what you’ll respond to, what you’ll ignore, and what will keep you engaged.

This model is not static. It updates with every interaction. And crucially, it doesn’t just reflect your behaviour. It shapes the conditions under which your behaviour occurs. If the AI decides you are a cautious, security-oriented customer and begins filtering your world accordingly, showing you safer options, presenting information with more caveats, emphasising guarantees over possibilities, does it make you more cautious? If it decides you are an adventurous early adopter and surfaces novelty at every turn, does it reinforce that orientation?

The answer, according to the psychological science that underpins the STAR Framework, is almost certainly yes. And the mechanism is well understood.

Self-Determination Theory: The Autonomy Question

Self-Determination Theory (SDT), one of the seven pillars of the STAR Framework, identifies three fundamental psychological needs: autonomy, competence, and relatedness. Of these three, autonomy is the one most directly threatened by algorithmic identity assignment.

Autonomy, in SDT terms, is not just about having choices. It’s about having a sense of volition, the feeling that your actions originate from within rather than being controlled by external forces. When you choose to be cautious because caution reflects your genuine orientation, that’s autonomous behaviour. When you become cautious because an algorithm has decided you are cautious and has structured your environment accordingly, the line between autonomous expression and algorithmic shaping becomes uncomfortably thin.

This is not a hypothetical concern. Every recommendation engine, every content filter, every personalisation algorithm is making assumptions about who you are and serving you a version of the world that matches those assumptions. The question SDT forces us to ask is: at what point does the algorithm’s model of your identity become a self-fulfilling prophecy?

The research suggests the answer is “sooner than you’d think.” Environmental framing, sustained patterns of choice architecture that nudge behaviour in a particular direction, can shift the expression of personality traits over time. Not the underlying traits themselves, which remain relatively stable. But the way those traits manifest in behaviour. The algorithm doesn’t change who you are. It changes who you become in the context it controls.

The Big Five: Stability and Plasticity

The Big Five personality model, another of STAR’s pillars, provides the empirical framework for understanding this dynamic. The Big Five identifies five broad dimensions of personality: openness, conscientiousness, extraversion, agreeableness, and neuroticism. These dimensions are relatively stable across the lifespan, which is why personality assessment is possible at all.

But “relatively stable” is not “immutable.” The Big Five research shows that personality traits can shift in response to sustained environmental pressures. A person who is moderately conscientious can become more or less so depending on the demands of their environment. A person who is moderately open to experience can become more or less so depending on the novelty of the stimuli they encounter.

This is where the feedback loop becomes genuinely concerning. If an AI system consistently serves a user with low-novelty, high-certainty content because it has classified them as a Realist-type, the sustained absence of novelty may reduce their openness to experience over time. The algorithm didn’t change their personality. It created an environment in which their personality expressed itself more narrowly.

Conversely, if an AI system consistently serves an Adventurer-type with high-novelty, high-stimulus content, it may reinforce their orientation towards novelty at the expense of other dimensions of their personality. The algorithm didn’t create the Adventurer. It amplified one.

The implication for the STAR Framework is significant. The twelve STAR archetypes are not fixed identities. They are dynamic patterns of motivational orientation that are shaped by context. If the context is increasingly mediated by AI systems that have already decided who you are, the archetypes become less descriptive and more prescriptive. The algorithm doesn’t just see the Adventurer. It builds the world the Adventurer inhabits.

Regulatory Focus Theory: Priming the Mindset

Regulatory Focus Theory, another pillar, adds another dimension. The theory distinguishes between promotion-focused individuals, who orient towards gains and opportunities, and prevention-focused individuals, who orient towards safety and the avoidance of loss. These orientations are not fixed traits. They are states that can be primed by environmental cues.

This is the mechanism through which algorithmic identity assignment operates. If an AI system determines that you are prevention-focused and begins filtering your world through that lens, emphasising risks, surfacing guarantees, and framing choices in terms of what you might lose, it is actively priming you into a prevention-focused state. The more consistently this framing occurs, the more entrenched the orientation becomes.

The reverse is equally true. A promotion-focused framing, one that emphasises opportunity, novelty, and potential gains, primes the user into a promotion-focused state. The algorithm isn’t just reading your regulatory focus. It’s reinforcing it.

This creates what I’d call a motivational gravity well. The algorithm identifies your orientation, builds a world that matches it, and the world it builds pulls you further into that orientation. The loop tightens. The range narrows. And the person you were, the one with a full spectrum of motivational possibilities, becomes increasingly defined by the single dimension the algorithm chose to amplify.

Social Identity Theory, another STAR pillar, describes how we derive our sense of self from the groups we belong to. We categorise the world into “us” and “them.” We identify with groups that reinforce our self-concept. We compare our group favourably to others.

AI systems are, in effect, creating digital tribes. Not formal groups with membership cards and meeting schedules, but clusters of users who share similar behavioural profiles and are served similar content. If the algorithm has decided you are a particular type, it will show you content that resonates with that type, connect you with perspectives that align with that type, and filter out perspectives that challenge it.

This is not conspiracy. It is optimisation. The algorithm’s goal is engagement, and engagement is maximised when the content feels relevant, familiar, and affirming. But the effect is tribal. You begin to see the world through the lens the algorithm has chosen for you, and that lens becomes increasingly familiar, increasingly comfortable, and increasingly difficult to see beyond.

Social Identity Theory predicts that this tribal framing will increase in-group identification and out-group dismissal. The algorithm doesn’t just show you content. It shows you your content, and in doing so, it reinforces the sense that the world the algorithm has built is the world as it actually is.

Cognitive Bias Theory: The Confirmation Engine

Cognitive Bias Theory, another pillar, explains why this feedback loop is so difficult to disrupt. The most relevant bias is confirmation bias, the tendency to seek, interpret, and remember information that confirms our existing beliefs. When an AI system serves you content that aligns with its model of who you are, confirmation bias ensures that you engage with that content more deeply, more frequently, and more approvingly than you would with content that challenges the model.

This is not a flaw in your reasoning. It is a feature of human cognition, one that evolved to help us process complex environments efficiently. But in an AI-mediated environment, confirmation bias becomes a confirmation engine. The algorithm serves content that confirms your profile. You engage with it. The algorithm interprets your engagement as validation. It serves more of the same. The loop accelerates.

The result is not just a filter bubble, which is a content problem. It is an identity bubble, which is a psychological problem. The algorithm doesn’t just limit what you see. It limits who you become by narrowing the range of experiences that might challenge, expand, or reshape your identity.

Appraisal Theory of Emotion: The Emotional Architecture

Appraisal Theory of Emotion, the seventh pillar, provides the final lens. Appraisal Theory describes how we evaluate events and situations for their relevance to our goals, needs, and wellbeing. The appraisal process is not conscious in most cases. It operates at the level of System 1 processing, fast, automatic, and largely invisible.

When an AI system serves you content, your emotional response is shaped by the appraisal process. If the content aligns with your goals and needs, you appraise it positively. If it challenges them, you appraise it negatively. The algorithm learns from these appraisals, optimising for positive emotional responses and filtering out negative ones.

This creates an emotional architecture that mirrors the motivational gravity well. The algorithm identifies the emotional patterns that drive your engagement, serves content that activates those patterns, and in doing so, reinforces the emotional architecture it has built. You don’t just see the world through the algorithm’s lens. You feel the world through it.

The Unoptimised Human

Here is the central dilemma, and it is the one that gave my book its title. In a world designed to optimise your behaviour, do you still get to choose who you are?

The STAR Framework says yes, but only if you understand the forces at play. The seven pillars that underpin STAR are not just tools for understanding other people. They are tools for understanding yourself, and, critically, for understanding the systems that are trying to understand you.

Self-Determination Theory tells you that autonomy is a need, not a luxury, and that environmental framing can erode it. The Big Five tells you that personality is stable but not immutable, and that sustained algorithmic shaping can narrow its expression. Regulatory Focus Theory tells you that your motivational orientation can be primed, and that the algorithm is priming it constantly. Social Identity Theory tells you that tribal identity is powerful, and that the algorithm is building your tribe for you. Cognitive Bias Theory tells you that confirmation bias makes the loop self-reinforcing. Appraisal Theory of Emotion tells you that the algorithm is not just shaping what you think but how you feel.

The defence is not to disengage from AI. That ship has sailed, and the benefits are real. The defence is to become aware of the loop. To recognise when the algorithm has assigned you a type and to consciously seek experiences that challenge the assignment. To ask, not “What does the algorithm think I want?” but “What would I choose if the algorithm weren’t choosing for me?”

This is what I mean by the unoptimised human. Not a person who rejects technology. Not a person who is immune to algorithmic influence. But a person who understands the influence well enough to maintain genuine agency within it.

The CMO’s Responsibility

For CMOs and marketing leaders, this has a direct and practical implication. The brands that thrive in an AI-mediated environment will not be the ones that best optimise for algorithmic engagement. They will be the ones that best understand the psychological dynamics of the feedback loop and build their presence accordingly.

This means creating content that doesn’t just speak to the algorithm’s model of the customer, but that speaks to the full range of the customer’s psychological possibilities. It means building brand experiences that challenge as well as affirm. It means recognising that the customer in front of the AI is not just the person the algorithm thinks they are, but the person they might become if given the chance.

The DOTS Framework, the operational extension of STAR, offers a practical starting point. Data for Thinkers. Opportunity for Adventurers. Togetherness for Socialisers. Stabilise for Realists. But the deeper principle is this: every customer is more than the profile the algorithm has built for them. The brands that honour that complexity will earn something that no algorithm can manufacture: genuine trust.

The Question That Remains

I titled my book Dear Algorithm, It’s Not Me, It’s You, and the joke was intentional. But the truth it points to is not a joke at all. The algorithm is constructing a version of you, and that version is becoming increasingly real, not because it’s accurate, but because the world is being built to match it.

The STAR Framework gives us the vocabulary to see this happening. The seven pillars give us the science to understand the mechanism. The twelve archetypes give us the map of the territory. But the choice, the one that matters, is yours.

Who are you when the algorithm isn’t watching?

The STAR Framework synthesises seven psychological theories into four primary mindsets and twelve distinct archetypes. It is the foundation of the STAR Operating System, an award-winning model for understanding human behaviour in organisational, consumer, and leadership contexts. David Chadderton is the creator of the STAR Framework and author of The STAR Framework: Rewriting the Rules of Consumer Engagement (NYC Big Book Award, 2025), The STAR Operating System, and Dear Algorithm, It’s Not Me, It’s You. He is CMO at Homes for Students, VervLife, and Orla.