AI Pricing Has a Fairness Problem

How consumer data can shape the price you see, and what to check before you pay.

You and another shopper could open the same delivery app for the same grocery order and see totals that differ by several dollars without any context.

The difference may come from timing, location, account status, fees, or how the platform estimates what you are likely to pay. Some price gaps may have ordinary explanations. Others may reflect a pricing system that knows more about you than you realize.

As consumers, we need better price transparency before we accept the price on the screen.

Dynamic pricing is not new. Airlines, hotels, rideshare platforms, and ticketing businesses have adjusted prices for years. What is different now is the data environment around pricing. AI and pricing algorithms can process more data, test more variations, and adjust prices or fees in ways that are harder for consumers to see.

That is why dynamic AI pricing is not only a business optimization story. It is a consumer fairness issue.

The problem is fairness. When AI helps decide what price you see, the same product can start to feel less like a market price and more like a personalized test of what you might tolerate. That is the part consumers should care about: whether the price change was fair, visible, and explainable.

Regulators are paying closer attention to pricing systems that use consumer data to shape what people pay. The Federal Trade Commission has described one version of this concern as surveillance pricing: using personal information to set targeted prices for goods and services. In its initial findings, the FTC said pricing intermediaries may use data such as location, demographics, browsing behavior, and even mouse movements on a webpage.

That is important because the price on the screen may reflect more than demand or inventory.

A grocery order, hotel room, flight, or rideshare fare can feel like a simple price on a screen. Behind that price may be a company trying to determine what a specific customer is likely to accept.

New York has also moved toward disclosure. Its algorithmic pricing law requires businesses to disclose when they use consumer-specific personal data in a pricing algorithm to set the price a New York consumer sees. That kind of notice is useful, but it does not answer the question most shoppers care about: am I seeing a fair price compared with someone else?

A label may tell you an algorithm is involved. It may not tell you whether another shopper saw a lower total, whether fees changed because of your location, or whether the platform is testing how much you will tolerate before you abandon the purchase.

Dynamic pricing and personalized pricing are related, but they are not the same thing.

Dynamic pricing usually responds to market conditions. Uber’s surge pricing is a familiar example: when many riders want a car in the same area at the same time, the fare can rise before you request the trip. The price is reacting to demand in that moment, not necessarily to who you are as an individual customer.

Personalized pricing goes further. It uses information about the person, account, device, behavior, location, or shopping history to influence the price or fee that person sees.

Surveillance pricing is the more sensitive concern because it points to a deeper data privacy concern. The price may be shaped by information gathered across digital activity, sometimes through companies the consumer never directly chose to deal with.

That distinction is important for fairness.

A higher price during peak demand is easier to understand. A higher price because a platform believes you are less likely to compare alternatives is different. One is the price reacting to demand. The other is the price reacting to what a system thinks it knows about you.

We do not need to treat every online price as suspicious. That would be exhausting and unrealistic.

A better habit is to compare the final totals before deciding which price is fair.

Start with the same item or order while logged in, then check it again while logged out or in a private browser. If the number changes, look closely at what changed: the base price, the fees, the promotion, or the final total.

That final total matters most. A platform can keep the item price unchanged while moving the real cost into fees. From the consumer’s perspective, the checkout total is the price.

It is also worth comparing the platform price with the direct seller price. This is especially useful for food delivery, travel, household goods, and marketplace purchases. The point is to avoid paying a higher total simply because the app made the comparison inconvenient.

Location and timing can matter too. A delivery address, pickup option, or ZIP code may change what appears on the screen. A price may also move during a demand spike or after repeated searches. Some of that may have ordinary explanations. The disparity becomes more important when similar shoppers, similar conditions, or the same account see meaningfully different totals.

Once you see a meaningful price gap, your next step is simple.

Before paying, choose the lower verified total. That may mean ordering directly from the seller instead of through the app, using the logged-out cart, or selecting the checkout screen with the lower final amount.

If you already paid, keep the receipt or final checkout total and ask customer support for a specific action: a price adjustment, a fee reversal, or a clear explanation of why the lower total was not available. The company may not be required to change the charge, but a precise request gives you a better chance of getting a useful answer.

Ordinary price movement may be frustrating, but it is not automatically deceptive.

Deceptive pricing is different. It can include advertising one price and adding mandatory fees later, presenting a discount that is not real, hiding important price terms until checkout, or using a misleading comparison. If the pattern looks deceptive, contact the company first. If the answer does not make sense, consumers can report the concern to the Federal Trade Commission or their state attorney general’s consumer protection office.

For companies, dynamic AI pricing can look attractive because it connects directly to revenue. If AI investment is expensive, pricing optimization can become one of the clearest ways to show a return. That connects to a broader issue I wrote about in The AI Spending Gap Is Starting to Show: companies spending heavily on AI will face pressure to turn capability into measurable business results.

Pricing may be one of the clearest ways to turn AI investment into revenue, but it can also create compliance risk when the company cannot explain who saw which price and why.

AI pricing systems can create legal, reputational, and operational risk when they produce disparities the business cannot explain. A third-party pricing tool may test price or fee changes without enough internal visibility. New York’s algorithmic pricing disclosure law can tell customers when consumer-specific personal data is being used to set a price. But a disclosure may satisfy a minimum requirement while leaving the real issue unresolved: customers may know an algorithm was used without knowing whether the price was fair, how their data affected it, or why someone else may have seen a lower total.

The governance work should be practical. Companies need to know what data is used, what prices or fees the system can change, who approved the pricing logic, how tests are documented, and how consumer complaints are handled. Legal, compliance, privacy, and business teams should understand the difference between ordinary dynamic pricing and individualized pricing based on personal data.

This is where AI governance and AI regulation meet the customer experience.

A pricing system can be technically sophisticated and still damage trust if customers believe the company is hiding the real price logic.

Dynamic AI pricing does not require us to distrust every online price. It does require a better consumer habit.

Before paying, ask a simple question: did the price change because the market changed, or because the system thinks I will accept it?

We may not always know the answer. But we can compare more often, check the final total, save proof when the difference matters, and question prices that do not make sense.

AI gives companies more ways to personalize the price in front of each customer.

Consumers need practical ways to recognize when that price deserves a second look.

AI generated images

Source block

• Federal Trade Commission: Surveillance Pricing
• Federal Trade Commission: Surveillance Pricing Study Indicates Wide Range of Personal Data Used to Set Individualized Consumer Prices
• New York State Senate: General Business Law Section 349-A, Algorithmic Pricing Disclosure
• Uber: How Surge Pricing Works
• Federal Trade Commission: Report Fraud
• Federal Trade Commission: Rule on Unfair or Deceptive Fees, Frequently Asked Questions

Originally published at https://icreview.substack.com.