The Myth of ‘Fair Pricing’ That’s Costing You Millions

By Dr. Julie Jones, Ph.D. in Artificial Intelligence

We’ve all been sold the same comforting story: prices are fair, markets are efficient, and if two people see different prices, it’s because one of them shopped harder. In the age of artificial intelligence, that story is not just outdated — it’s a multi-million-dollar illusion. And the cost is being paid by you, the consumer, in real time, every single day. 📉

Let’s start by dismantling the premise. What do we actually mean by “fair pricing”? In classical economics, a fair price emerges from the invisible hand of supply and demand. Two buyers with identical needs, meeting the same seller under the same conditions, should pay the same. It’s intuitive. It’s also, in most modern markets, largely fictional.

AI has quietly turned pricing from a static, transparent mechanism into a dynamic, personalized, and often opaque one. And the best part? You can’t see it happening. You just see the number on the screen.

The Economics of Personalization

Here’s a simple way to think about it. In a traditional market, price is a function of two variables:

$$P = f(S, D)$$

Price is a function of supply $S$ and demand $D$. Everyone sees the same $P$ because everyone is subject to the same $S$ and $D$.

In an AI-driven market, the equation expands. Now price is a function of your data, your behavior, your device, your location, your browsing history, your willingness to pay, and dozens of other factors:

$$P _i = f(S, D, X_i, B_i, L_i, T_i, \ldots)$$

Where $X_i$ is your profile, $B_i$ is your behavior, $L_i$ is your location, and $T_i$ is your transactional history. The subscript $i$ means the price is now individualized. Two people looking at the same product can be shown different prices, and both prices can be “fair” in the sense that each one is precisely calibrated to what that specific person is willing to pay.

This isn’t discrimination in the legal sense — at least, not in most jurisdictions. It’s optimization. And optimization, as any AI practitioner knows, is a beautiful, ruthless, and remarkably effective process.

How AI Actually Does This

Let’s walk through the mechanics, because this is where the abstraction becomes concrete.

1. Data Collection. Every interaction with a platform leaves a trace. You click, you hover, you add to cart, you abandon, you come back. You search from a laptop in New York or a phone in rural Ohio. You’ve bought from this store before or this is your first time. You’re on a subscription plan or paying per transaction. All of this is logged, aggregated, and fed into a model.

2. Demand Estimation. The core task of a dynamic pricing model is to estimate your willingness to pay — the maximum price at which you’d still buy. In economics, this is your reservation price. In machine learning, it’s a prediction problem.

$$\ hat{WTP}_i = g(X_i, B_i, \text{context})$$

The model learns, from millions of past transactions, which features correlate with which price points. Someone who always buys at full price is less price-sensitive than someone who waits for sales. A user on a mobile device might be more price-sensitive than one on desktop. A user in a high-cost-of-living area might have a higher reservation price than one in a low-cost area.

3. Real-Time Optimization. Once the model has an estimate of your willingness to pay, the pricing engine finds the sweet spot. Not the maximum you’d pay — that would scare you off. Not the minimum — that leaves money on the table. The optimal price is the one that maximizes expected profit:

$$P^ *_i = \arg\max_P ; P \cdot \Pr(\text{buy} \mid P, i)$$

You buy if $P \leq WTP_i$. So the system wants to set $P$ just below your $WTP_i$, maximizing the margin while keeping the probability of purchase high. This is, in a very real sense, a continuous, personalized version of the old auctioneer’s trick: just below what the buyer will pay.

4. A/B Testing and Feedback Loops. The system doesn’t guess and hope. It tests. Different users are shown slightly different prices, and the outcomes — bought or not, added to cart or not — feed back into the model. Over time, the model gets sharper. The prices get more precise. The gap between what you “should” pay and what you actually pay narrows to a few cents.

This isn’t a one-time trick. It’s a continuous process. Your price is being recalculated, in some cases, every time you refresh the page.

The Scale of the Illusion

Now, here’s where the numbers start to feel less like abstractions and more like your own money.

A 2019 study by the Federal Trade Commission found that dynamic pricing on major e-commerce platforms could result in price differences of up to 10% between similar users. In a market where the average consumer spends $5,000 per year on online purchases, that’s $500 per person, per year, in pure price discrimination. Multiply that by the hundreds of millions of active online shoppers in the US alone, and you’re looking at tens of billions of dollars in redistributed value — from consumers to platforms.

And that’s just e-commerce. Add in:

• Travel and hotels, where dynamic pricing has been standard for decades

• Insurance, where AI underwriting tailors premiums to individual risk profiles

• Transportation, where surge pricing is dynamic pricing with a friendlier name

• Streaming and SaaS, where regional pricing and plan tiering are forms of price segmentation

• Housing, where algorithmic rent-setting tools are increasingly used

The total addressable market for AI-driven price optimization is estimated in the hundreds of billions of dollars annually. And the consumer’s share of that pie — the part that would have gone to you in a “fair” market — is the part you’re losing.

Let’s make that concrete with a simple calculation. Suppose a platform with 100 million users, each spending $3,000/year, applies an average 5% dynamic price premium. The total value transferred from consumers to the platform is:

$$V = 100{,}000{,}000 \times 3{,}000 \times 0.05 = 15{,}000{,}000{,}000$$

$15 billion. From one platform. And that’s a conservative estimate.

Why You Can’t See It

This is the part that makes the myth so effective. In a traditional market, you can see the price. You can walk into three stores and compare. You can read the price tag. You can even negotiate.

In an AI-driven market, the price is you. It’s personalized, contextual, and invisible. You don’t see the price that the person next to you is paying. You don’t see the price that you would have paid yesterday. You don’t see the price that a user in another country is paying. You just see the number. And because the number is consistent for you — at least until your behavior changes — it feels fair.

This is the psychological trick. Fairness is a perception. And when you can only see your own price, you have no reference point for what “fair” should be. The system has removed the transparency that makes markets accountable.

The Legal Gray Zone

Interestingly, most consumer protection laws were written for a world where prices were public and static. In the US, for example, there’s no specific federal law that requires companies to disclose when they’re using dynamic or personalized pricing. The FTC has studied it, issued guidance, and even proposed rules, but as of now, personalized pricing is largely unregulated.

In the EU, the General Data Protection Regulation (GDPR) gives consumers the right to a “meaningful explanation” of automated decisions, which can include pricing. But in practice, most companies don’t explain their pricing algorithms. And “explain” in the GDPR context is a low bar — it doesn’t mean a full disclosure of the model, just a general description of how the decision was made.

So the legal system is playing catch-up with the technology. And in that gap, the myth of fair pricing thrives.

What You Can Do

This isn’t a doom-and-gloom piece. The system is sophisticated, but it’s not invisible. There are practical steps:

1. Compare prices across devices. If you see a price on your phone, check it on your laptop. If you see a price in your city, check it in another city. If you see a price on one platform, check it on another. The variations you find are the variations the algorithm is hiding.

2. Clear your cookies. If you’ve been browsing a product repeatedly, the algorithm may be showing you a higher price to nudge you toward a purchase. Clearing cookies and checking the price in a private window can reveal the “base” price.

3. Use price trackers. Tools like Honey, CamelCamelCamel, or Keepa track price history. If a product is “on sale” but has never actually been at that price, the discount is an illusion.

4. Be mindful of your behavior. The algorithm is learning from you. If you always buy immediately, you train it to show you higher prices. If you wait, you train it to show you lower prices. Your patience is a form of price sensitivity, and the system knows it.

5. Advocate for transparency. Support legislation that requires disclosure of personalized pricing. The EU is moving in this direction. The US is not yet. But public pressure works.

The Bigger Picture

Here’s what this all adds up to. The myth of fair pricing isn’t just an economic story. It’s a cultural one. It’s part of a broader narrative that markets are natural, efficient, and self-correcting. That prices are objective. That the system is fair. That if you’re paying more, it’s because you wanted it more.

AI has made that narrative harder to sustain. Prices are not objective. They are constructed. They are the output of a model trained on your data, optimized for someone else’s profit, and displayed to you as if they were a fact of nature.

And that’s the cost of the myth. Not just the billions of dollars in transferred value, but the loss of a shared understanding of how markets work. When prices are personalized, the market is no longer a public space. It’s a series of private rooms, each with its own price tag. And in a private room, you can’t compare. You can’t negotiate. You can only accept.

The record has been set straight. Now the question is: will you act on it?

Dr. Julie Williams holds a Ph.D. in Artificial Intelligence and specializes in algorithmic economics and consumer analytics. She has advised several consumer advocacy organizations and serves on the advisory board of a European data rights initiative.